From 1fbe7c39d4cf934ecb5315cfa51eddb390340d30 Mon Sep 17 00:00:00 2001 From: lintianle Date: Tue, 7 Jul 2026 23:21:54 +0800 Subject: [PATCH 01/26] feat: add MCP client plugin (dsh-mcp-client) Connects to an external MCP server and registers its tools on ctx.tools. Supports stdio (child process) and Streamable HTTP transports. Credential-shaped env vars are scrubbed before forwarding to child processes. - Plugin lifecycle: connect, sync tools, re-sync on ToolListChanged, dispose unregisters and closes - Full JSDoc on all exports (@param/@returns on functions) - 100% per-file coverage (apply lifecycle, args coercion, env scrubbing) - Config catalog regenerated --- docs/config-catalog.md | 43 ++ docs/module-graph.md | 6 + docs/rfc/INDEX.md | 1 + .../feature/2026-07-07-mcp-client-plugin.md | 160 ++++ packages/mcp/README.md | 7 + packages/mcp/mcp-client/README.md | 57 ++ packages/mcp/mcp-client/package.json | 38 + packages/mcp/mcp-client/src/index.ts | 128 ++++ packages/mcp/mcp-client/src/tools.ts | 168 +++++ packages/mcp/mcp-client/src/transport.ts | 56 ++ packages/mcp/mcp-client/tests/apply.spec.ts | 179 +++++ .../mcp/mcp-client/tests/mcp-client.spec.ts | 518 +++++++++++++ packages/mcp/mcp-client/tsconfig.json | 15 + pnpm-lock.yaml | 701 +++++++++++++++++- scripts/gen-config-catalog.ts | 9 + tsconfig.base.json | 1 + tsconfig.build.json | 3 +- tsconfig.json | 3 +- 18 files changed, 2087 insertions(+), 6 deletions(-) create mode 100644 docs/rfc/proposed/feature/2026-07-07-mcp-client-plugin.md create mode 100644 packages/mcp/README.md create mode 100644 packages/mcp/mcp-client/README.md create mode 100644 packages/mcp/mcp-client/package.json create mode 100644 packages/mcp/mcp-client/src/index.ts create mode 100644 packages/mcp/mcp-client/src/tools.ts create mode 100644 packages/mcp/mcp-client/src/transport.ts create mode 100644 packages/mcp/mcp-client/tests/apply.spec.ts create mode 100644 packages/mcp/mcp-client/tests/mcp-client.spec.ts create mode 100644 packages/mcp/mcp-client/tsconfig.json diff --git a/docs/config-catalog.md b/docs/config-catalog.md index 31cff65d9b..2a99f439e6 100644 --- a/docs/config-catalog.md +++ b/docs/config-catalog.md @@ -353,6 +353,49 @@ export interface Config { Source: [`packages/support/llm-replay/src/index.ts:429`](../packages/support/llm-replay/src/index.ts) +## `@deepseek-ai/dsh-mcp-client` + +Requires: `tools` + +```ts config-catalog +/** Discriminated union of all supported MCP transport configurations. */ +export type Config = StdioConfig | StreamableHttpConfig + +/** Config for connecting to an MCP server via a spawned child process over stdio. */ +export interface StdioConfig { + /** Transport type: spawn a child process and communicate over stdio. */ + transport: 'stdio' + /** Executable to spawn. */ + command: string + /** Arguments passed to the command. */ + args: string[] + /** Extra env vars merged on top of scrubbed ambient env. */ + env: Record + /** Working directory for the child process. */ + cwd: string + /** Prefix prepended to each tool name before registration. */ + toolPrefix: string + /** Timeout per callTool invocation (ms). */ + toolCallTimeoutMs: number +} + +/** Config for connecting to an MCP server over Streamable HTTP (SSE). */ +export interface StreamableHttpConfig { + /** Transport type: connect to an MCP server over Streamable HTTP (SSE). */ + transport: 'streamable-http' + /** MCP server URL. */ + url: string + /** Extra headers (e.g. auth tokens). */ + headers: Record + /** Prefix prepended to each tool name before registration. */ + toolPrefix: string + /** Timeout per callTool invocation (ms). */ + toolCallTimeoutMs: number +} +``` + +Source: [`packages/mcp/mcp-client/src/index.ts:66`](../packages/mcp/mcp-client/src/index.ts) + ## `@deepseek-ai/dsh-session-persistence-jsonl` Requires: `sessions` diff --git a/docs/module-graph.md b/docs/module-graph.md index ed1cc592d4..05acb188d9 100644 --- a/docs/module-graph.md +++ b/docs/module-graph.md @@ -78,6 +78,9 @@ flowchart TD pkg_app_boot["app-boot"] pkg_stdio_agent["stdio-agent"] end + subgraph group_mcp["packages/mcp"] + pkg_mcp_client["mcp-client"] + end pkg_llm --> pkg_brand pkg_bash --> pkg_brand pkg_llm_deepseek --> pkg_llm @@ -156,6 +159,8 @@ flowchart TD pkg_acp --> pkg_session pkg_acp --> pkg_session_persistence pkg_acp --> pkg_tools + pkg_mcp_client --> pkg_llm + pkg_mcp_client --> pkg_tools pkg_agent_core --> pkg_agent pkg_agent_core --> pkg_agent_loop pkg_agent_core --> pkg_invariants @@ -239,6 +244,7 @@ flowchart TD | [`tool-todo`](../packages/todo/tool-todo) | `todo` | [`agent`](../packages/core/agent), [`session`](../packages/core/session), [`tools`](../packages/core/tools) | | [`hooks-codex`](../packages/hooks/hooks-codex) | `hooks` | [`agent`](../packages/core/agent), [`hook-protocol`](../packages/hooks/hook-protocol), [`llm`](../packages/llm/llm), [`session`](../packages/core/session), [`tools`](../packages/core/tools) | | [`acp`](../packages/ui/acp) | `ui` | [`agent`](../packages/core/agent), [`llm`](../packages/llm/llm), [`session`](../packages/core/session), [`session-persistence`](../packages/session-persistence/session-persistence), [`tools`](../packages/core/tools) | +| [`mcp-client`](../packages/mcp/mcp-client) | `mcp` | [`llm`](../packages/llm/llm), [`tools`](../packages/core/tools) | | [`agent-core`](../packages/core/agent-core) | `core` | [`agent`](../packages/core/agent), [`agent-loop`](../packages/core/agent-loop), [`invariants`](../packages/support/invariants), [`llm`](../packages/llm/llm), [`session`](../packages/core/session), [`system-prompt`](../packages/core/system-prompt), [`tool-bash`](../packages/bash/tool-bash), [`tools`](../packages/core/tools) | | [`subagent-acp`](../packages/subagent/subagent-acp) | `subagent` | [`agent`](../packages/core/agent), [`llm`](../packages/llm/llm), [`subagent`](../packages/subagent/subagent) | | [`subagent-inprocess`](../packages/subagent/subagent-inprocess) | `subagent` | [`agent`](../packages/core/agent), [`llm`](../packages/llm/llm), [`session`](../packages/core/session), [`subagent`](../packages/subagent/subagent) | diff --git a/docs/rfc/INDEX.md b/docs/rfc/INDEX.md index d69aaf87db..3483919692 100644 --- a/docs/rfc/INDEX.md +++ b/docs/rfc/INDEX.md @@ -12,6 +12,7 @@ Generated by `pnpm run gen-rfc-index` from the RFC tree — never edit by hand; | [Multiplex concurrent ACP sessions over one connection](proposed/feature/2026-06-14-acp-multi-session.md) | 2026-06-14 | | [Optional Code Mode — model writes TypeScript against an SDK of all tools](proposed/feature/2026-06-15-optional-code-mode.md) | 2026-06-15 | | [Pre-tool input rewrite — a consistent design](proposed/feature/2026-06-30-pre-tool-input-rewrite.md) | 2026-06-30 | +| [MCP client plugin — connect to external MCP servers and bridge their tools](proposed/feature/2026-07-07-mcp-client-plugin.md) | 2026-07-07 | ### Simplification diff --git a/docs/rfc/proposed/feature/2026-07-07-mcp-client-plugin.md b/docs/rfc/proposed/feature/2026-07-07-mcp-client-plugin.md new file mode 100644 index 0000000000..952335f840 --- /dev/null +++ b/docs/rfc/proposed/feature/2026-07-07-mcp-client-plugin.md @@ -0,0 +1,160 @@ +# RFC: MCP client plugin — connect to external MCP servers and bridge their tools + +Status: proposed + +## Problem + +The harness has no way to consume tools from the MCP (Model Context Protocol) ecosystem. MCP is the emerging standard for tool servers — GitHub, filesystem, databases, code search, and hundreds of community servers expose tools via MCP. Users want to point the harness at one or more MCP servers and have their tools appear as native model-facing tools, without writing per-server glue code. + +The `ToolRegistry` already accepts raw JSON Schema tool definitions (documented in `dsh-tools` README: "Raw JSON-Schema tool definitions (from MCP servers) are still accepted by `ToolRegistry.register()` directly"), and the extension cookbook sketches the intended pattern ("MCP | one plugin per server: discover tools → `ctx.tools.register()`"). The infrastructure is ready; the bridge plugin is missing. + +## Proposal + +### Package + +A single package `@deepseek-ai/dsh-mcp-client` at `packages/mcp/mcp-client/`. No capability-seam three-package split — there is no foreseeable second MCP client implementation, and the convention is "don't split preemptively" ([capability seams RFC](../../implemented/architecture/2026-06-13-capability-seams.md)). + +### SDK + +Use the official [`@modelcontextprotocol/sdk`](https://github.com/modelcontextprotocol/typescript-sdk) (`Client`, `StdioClientTransport`, `StreamableHTTPClientTransport`). The harness does not implement its own JSON-RPC — consistent with how ACP delegates to `@agentclientprotocol/sdk`. + +### Scope + +MCP Client only (no server side — ACP already covers the "expose harness as an agent" role). Bridge **Tools** only — Resources and Prompts are deferred (they require harness-side consumption mechanisms that don't exist yet, and design space is large). + +### Plugin shape + +Namespace plugin (named exports `name`/`inject`/`Config`/`apply`, no `export default`). `inject: ['tools']`. Each MCP server is one plugin instance in `cordis.yml` — the same package loaded N times with different configs, like `dsh-tool-subagent`. + +### Configuration + +Flat discriminated union on the `transport` field: + +```typescript +interface StdioConfig { + transport: 'stdio' + command: string + args?: string[] + env?: Record + cwd?: string + toolPrefix?: string + toolCallTimeoutMs?: number // default 60_000 +} + +interface StreamableHttpConfig { + transport: 'streamable-http' + url: string + headers?: Record + toolPrefix?: string + toolCallTimeoutMs?: number // default 60_000 +} + +type Config = StdioConfig | StreamableHttpConfig +``` + +Example `cordis.yml` usage: + +```yaml +- id: mcp-github + name: '@deepseek-ai/dsh-mcp-client' + config: + transport: stdio + command: npx + args: ['-y', '@modelcontextprotocol/server-github'] + env: + GITHUB_TOKEN: !!js process.env.GITHUB_TOKEN + +- id: mcp-web + name: '@deepseek-ai/dsh-mcp-client' + config: + transport: streamable-http + url: http://localhost:3000/mcp + headers: + Authorization: !!js `Bearer ${process.env.MCP_TOKEN}` +``` + +### Lifecycle + +Boot-time from `cordis.yml`. HMR (`@cordisjs/plugin-hmr`) provides hot-swap: editing the yml entry triggers dispose of the old instance (disconnects, unregisters tools) and creation of a new one (connects, discovers, registers). No runtime-dynamic API for now. + +### Tool discovery and registration + +1. On connect: `client.listTools()` → register each tool as a raw `ToolDefinition` via `ctx.tools.register()`. +2. Listen for `notifications/tools/list_changed` → re-run `listTools()`, diff, unregister removed, register added. +3. Registration uses the raw JSON Schema from MCP (no `defineTool` DSL conversion). +4. No `presentCall`/`presentResult` — the ACP bridge's generic-card fallback handles rendering. +5. Tools are transparent in the system prompt — no "[via MCP]" annotation. + +### Name conflict handling + +If `config.toolPrefix` is set (e.g. `"gh_"`), it is prepended to each MCP tool name before registration. If a name collides with an already-registered tool, log a warning and skip that tool (do not crash the entire server connection). + +### Tool execution + +A unified `execute` handler for all tools from one MCP server: + +1. Call `client.callTool({ name, arguments }, { signal: exec.signal })` with the configured timeout. +2. Map the result: + - Multiple `text` content blocks → join with `'\n'` into a single `TextBlock` (required: `flattenText` uses `join('')` without separator, so multiple blocks would lose inter-block boundaries). + - `image` content blocks → discard with a `ctx.logger.warn` (the harness has no image content block type; [drop-image RFC](../../implemented/simplification/2026-07-04-drop-image-content-block.md)). + - `isError: true` → map to the harness `isError` result path (`{ content: [...], isError: true }`). +3. Cancellation: `exec.signal` (from the agent loop's cancel) is passed through to the MCP SDK's `callTool`, which sends `$/cancelRequest` to the server. + +### Subprocess environment (stdio transport) + +Replicate the `buildChildEnv` + `SENSITIVE_ENV_PATTERN` scrub from `dsh-subagent-acp`: filter ambient env (strip credential-shaped vars matching `/KEY|SECRET|TOKEN/i`), then merge `config.env` on top. Explicit env overrides survive the scrub. + +### Disconnection / crash + +No auto-reconnect. If the MCP server process exits or the transport closes: + +1. The effect disposes → all registered tools are unregistered (fiber-scoped disposers). +2. Subsequent model calls to those tools → `ToolNotFoundError` → `isError: true`. +3. Recovery: user edits `cordis.yml` (triggers HMR reload) or restarts the harness. + +This matches the ACP subagent pattern: "crash = terminal, report error, clean up, don't retry." + +## Alternatives considered + +### MCP Server side (expose harness tools to external MCP clients) + +Deferred. The ACP bridge already exposes the harness as an agent server. Adding an MCP server layer would duplicate that with a different protocol, and the primary user need is consuming external tools, not exposing them. + +### Capability-seam three-package split (interface / impl / consumer) + +Rejected. There is no foreseeable alternative MCP client implementation — MCP has one protocol, one SDK. The convention is "don't split preemptively" until a second implementation appears. + +### Auto-reconnect with exponential backoff + +Rejected for v1. Adds complexity (partial-availability state where tools are registered but temporarily non-functional), and stdio process crashes usually indicate a configuration problem that retrying won't fix. HMR already provides the manual recovery path. Can be added as a future `reconnect: boolean` config if needed. + +### Bridge Resources and Prompts + +Deferred. Resources need a harness-side mechanism to decide WHEN to inject content (system prompt? on demand? model-triggered?). Prompts need a "prompt template" concept the harness lacks. Both require their own design; Tools are the high-value, low-risk starting point. + +### Always-on namespace prefix (e.g. `mcp_github__create_issue`) + +Rejected. Most MCP servers already use semantic prefixes in their tool names (e.g. `github_create_issue`). A forced prefix would break model familiarity with well-known MCP tool names and waste context tokens. Optional `toolPrefix` handles the rare collision case. + +### Preserve multiple TextBlocks in tool result + +Rejected. `flattenText()` in the DeepSeek serializer uses `join('')` (no separator) when flattening `ContentBlock[]` to wire format. Multiple text blocks would silently lose inter-block boundaries — a correctness bug. All existing tools return a single TextBlock; the MCP bridge follows suit. + +## Acceptance criteria + +- A `cordis.yml` entry connecting to an MCP stdio server (e.g. `@modelcontextprotocol/server-filesystem`) results in that server's tools appearing in the model's tool list and being callable. +- A `cordis.yml` entry connecting via Streamable HTTP works equivalently. +- Adding/removing an MCP entry in `cordis.yml` while HMR is active hot-swaps the tools without restart. +- `toolPrefix` config correctly prefixes tool names; a name collision logs a warning and skips. +- Agent cancel propagates to in-flight `callTool` (abort signal). +- Timeout fires and produces an `isError` result when an MCP server hangs. +- Server crash cleanly unregisters tools (no orphaned tool definitions). +- `notifications/tools/list_changed` triggers a re-sync of tool registrations. +- 100% test coverage on the new package (unit tests with mocked MCP SDK). + +## Risks + +- **MCP SDK stability**: the `@modelcontextprotocol/sdk` is still evolving. Breaking changes in the SDK require updating the bridge. Mitigation: pin a specific version; the SDK is widely adopted (Claude Desktop, Cursor, VS Code) so breaking changes are unlikely to be silent. +- **Tool schema quality**: MCP servers may expose poorly-described tools (vague descriptions, incomplete JSON schemas). The harness passes them through as-is — garbage-in-garbage-out. Mitigation: this is the server author's responsibility, not the bridge's. +- **Stdio process management**: a misbehaving MCP server that ignores signals could wedge dispose. Mitigation: the Cordis fiber disposal has bounded quiescence; a stuck transport eventually times out at the framework level. +- **Token budget pressure**: connecting many MCP servers with many tools inflates the system prompt. Mitigation: no different from registering many native tools; the compaction layer handles context pressure. diff --git a/packages/mcp/README.md b/packages/mcp/README.md new file mode 100644 index 0000000000..153afde8a9 --- /dev/null +++ b/packages/mcp/README.md @@ -0,0 +1,7 @@ +# MCP — Model Context Protocol + +Packages bridging the harness to the MCP ecosystem. + +| Package | Role | +|---|---| +| `mcp-client/` | MCP client bridge: connects to external MCP servers and registers their tools on `ctx.tools` | diff --git a/packages/mcp/mcp-client/README.md b/packages/mcp/mcp-client/README.md new file mode 100644 index 0000000000..68f6ebf42a --- /dev/null +++ b/packages/mcp/mcp-client/README.md @@ -0,0 +1,57 @@ +# @deepseek-ai/dsh-mcp-client + +MCP client bridge plugin: connects to external [Model Context Protocol](https://modelcontextprotocol.io/) servers and registers their tools on `ctx.tools`, making them available to the model as native tools. + +## Usage + +One plugin instance per MCP server in `cordis.yml`: + +```yaml +- id: mcp-github + name: '@deepseek-ai/dsh-mcp-client' + config: + transport: stdio + command: npx + args: ['-y', '@modelcontextprotocol/server-github'] + env: + GITHUB_TOKEN: !!js process.env.GITHUB_TOKEN + +- id: mcp-web + name: '@deepseek-ai/dsh-mcp-client' + config: + transport: streamable-http + url: http://localhost:3000/mcp + headers: + Authorization: !!js '`Bearer ${process.env.MCP_TOKEN}`' +``` + +HMR hot-swaps: editing the entry triggers disconnect + reconnect without process restart. + +## Config + +| Field | Transport | Required | Description | +|---|---|---|---| +| `transport` | both | yes | `"stdio"` or `"streamable-http"` | +| `command` | stdio | yes | Executable to spawn | +| `args` | stdio | no | Arguments passed to the command | +| `env` | stdio | no | Extra env vars merged on top of scrubbed ambient env | +| `cwd` | stdio | no | Working directory for the child process | +| `url` | http | yes | MCP server URL | +| `headers` | http | no | Extra headers (e.g. auth tokens) | +| `toolPrefix` | both | no | Prefix prepended to each tool name before registration | +| `toolCallTimeoutMs` | both | no | Timeout per `callTool` invocation (default 60000) | + +## Behavior + +- On connect: `listTools()` → registers each tool via `ctx.tools.register()`. +- Listens for `notifications/tools/list_changed` → re-syncs tool registrations. +- Tool execute: `client.callTool({ name, arguments }, { signal })` with timeout + abort support. +- Image content in results is discarded with a warning (the harness has no image block type). +- On disconnect/crash: all tools are unregistered; no auto-reconnect. +- Name conflicts: if a tool name collides, it is skipped with a warning. Use `toolPrefix` to disambiguate. + +## Services consumed + +| Service | Usage | +|---|---| +| `ctx.tools` | Register/unregister MCP tools | diff --git a/packages/mcp/mcp-client/package.json b/packages/mcp/mcp-client/package.json new file mode 100644 index 0000000000..69626cc606 --- /dev/null +++ b/packages/mcp/mcp-client/package.json @@ -0,0 +1,38 @@ +{ + "name": "@deepseek-ai/dsh-mcp-client", + "description": "MCP client bridge: connects to MCP servers and registers their tools on ctx.tools", + "version": "0.0.1", + "private": true, + "type": "module", + "main": "lib/index.js", + "types": "lib/types/index.d.ts", + "exports": { + ".": { + "types": "./lib/types/index.d.ts", + "default": "./lib/index.js" + }, + "./src/*": "./src/*", + "./package.json": "./package.json" + }, + "files": [ + "lib/index.js", + "lib/types/**/*.d.ts", + "lib/types/**/*.d.ts.map", + "src" + ], + "license": "BSD-3-Clause", + "peerDependencies": { + "@deepseek-ai/dsh-tools": "^0.0.1", + "@deepseek-ai/dsh-llm": "^0.0.1", + "cordis": "^4.0.0-rc.6" + }, + "dependencies": { + "@modelcontextprotocol/sdk": "^1.12.0", + "schemastery": "^3.18.0" + }, + "devDependencies": { + "@deepseek-ai/dsh-tools": "workspace:^", + "@deepseek-ai/dsh-llm": "workspace:^", + "cordis": "^4.0.0-rc.6" + } +} diff --git a/packages/mcp/mcp-client/src/index.ts b/packages/mcp/mcp-client/src/index.ts new file mode 100644 index 0000000000..da3aefc3de --- /dev/null +++ b/packages/mcp/mcp-client/src/index.ts @@ -0,0 +1,128 @@ +/** + * MCP client bridge plugin: connects to an external MCP server and registers + * its tools on `ctx.tools`. Each plugin instance connects to one MCP server; + * load multiple instances in `cordis.yml` for multiple servers. + * + * Namespace plugin (named exports, no default export). Lifecycle is + * effect-scoped: disposal disconnects from the server and unregisters all + * tools. HMR hot-swaps by disposing the old instance and creating a new one. + * + * @module @deepseek-ai/dsh-mcp-client + */ + +import type { Context } from 'cordis' +import z from 'schemastery' +import { Client } from '@modelcontextprotocol/sdk/client/index.js' +import { ToolListChangedNotificationSchema } from '@modelcontextprotocol/sdk/types.js' +import { createTransport } from './transport.ts' +import { syncTools } from './tools.ts' +// Side-effect type import: declaration-merges `ctx.tools` onto Context. +import type {} from '@deepseek-ai/dsh-tools' + +/** Cordis plugin name used by loader diagnostics. */ +export const name = 'mcp-client' + +/** Services required by this plugin. */ +export const inject = ['tools'] + +/** Default timeout for individual MCP tool calls (ms). */ +const DEFAULT_TOOL_CALL_TIMEOUT_MS = 60_000 + +// ---- Config ---- + +/** Config for connecting to an MCP server via a spawned child process over stdio. */ +export interface StdioConfig { + /** Transport type: spawn a child process and communicate over stdio. */ + transport: 'stdio' + /** Executable to spawn. */ + command: string + /** Arguments passed to the command. */ + args: string[] + /** Extra env vars merged on top of scrubbed ambient env. */ + env: Record + /** Working directory for the child process. */ + cwd: string + /** Prefix prepended to each tool name before registration. */ + toolPrefix: string + /** Timeout per callTool invocation (ms). */ + toolCallTimeoutMs: number +} + +/** Config for connecting to an MCP server over Streamable HTTP (SSE). */ +export interface StreamableHttpConfig { + /** Transport type: connect to an MCP server over Streamable HTTP (SSE). */ + transport: 'streamable-http' + /** MCP server URL. */ + url: string + /** Extra headers (e.g. auth tokens). */ + headers: Record + /** Prefix prepended to each tool name before registration. */ + toolPrefix: string + /** Timeout per callTool invocation (ms). */ + toolCallTimeoutMs: number +} + +/** Discriminated union of all supported MCP transport configurations. */ +export type Config = StdioConfig | StreamableHttpConfig + +export const Config = z.union([ + z.object({ + transport: z.const('stdio'), + command: z.string().required(), + args: z.array(String).default([]), + env: z.dict(String).default({}), + cwd: z.string().default(''), + toolPrefix: z.string().default(''), + toolCallTimeoutMs: z.number().default(DEFAULT_TOOL_CALL_TIMEOUT_MS), + }), + z.object({ + transport: z.const('streamable-http'), + url: z.string().required(), + headers: z.dict(String).default({}), + toolPrefix: z.string().default(''), + toolCallTimeoutMs: z.number().default(DEFAULT_TOOL_CALL_TIMEOUT_MS), + }), +]) as unknown as z + +// ---- Plugin apply ---- + +export function apply(ctx: Context, config: Config): void { + const transport = createTransport(config) + const client = new Client( + { name: 'dsh-mcp-client', version: '0.0.1' }, + { capabilities: {} }, + ) + + // Connect and set up tools. Errors during connect are logged, not thrown + // (the plugin simply has no tools registered). + const ready = (async () => { + await client.connect(transport) + + let disposers = await syncTools(client, ctx, { + toolPrefix: config.toolPrefix, + toolCallTimeoutMs: config.toolCallTimeoutMs, + }, new Map()) + + client.setNotificationHandler( + ToolListChangedNotificationSchema, + async () => { + ctx.logger.info('mcp-client: tool list changed, re-syncing') + disposers = await syncTools(client, ctx, { + toolPrefix: config.toolPrefix, + toolCallTimeoutMs: config.toolCallTimeoutMs, + }, disposers) + }, + ) + + return disposers + })().catch((error: unknown) => { + ctx.logger.error(`mcp-client: failed to connect: ${String(error)}`) + return new Map void>() + }) + + ctx.effect(() => async () => { + const disposers = await ready + for (const dispose of disposers.values()) dispose() + try { await client.close() } catch { /* transport already gone */ } + }, 'mcp-client.connection') +} diff --git a/packages/mcp/mcp-client/src/tools.ts b/packages/mcp/mcp-client/src/tools.ts new file mode 100644 index 0000000000..c3a35ccfba --- /dev/null +++ b/packages/mcp/mcp-client/src/tools.ts @@ -0,0 +1,168 @@ +/** + * Tool bridge: discovers MCP tools, registers them on the harness ToolRegistry, + * and handles re-sync when the server's tool list changes. + * + * @module + */ + +import type { Client } from '@modelcontextprotocol/sdk/client/index.js' +import type { Context } from 'cordis' +import type { ToolDefinition, ToolExecution } from '@deepseek-ai/dsh-tools' + +/** Resolved options relevant to tool bridging. */ +export interface ToolBridgeOptions { + toolPrefix: string + toolCallTimeoutMs: number +} + +/** State for one sync generation: the current set of disposers keyed by tool name. */ +type ToolDisposers = Map void> + +/** + * Sync the MCP server's tool list into the harness ToolRegistry. + * + * - Calls `client.listTools()` (paginated: drains all pages). + * - Registers each tool as a raw `ToolDefinition`. + * - On name conflict: logs a warning and skips that tool. + * - Returns a disposer map; call each value to unregister. + * + * @param client - Connected MCP Client instance used to list and call tools. + * @param ctx - Cordis context providing the `tools` service for registration. + * @param opts - Bridge options: tool name prefix and per-call timeout. + * @param previous - Disposer map from a prior sync generation; all entries are + * disposed before re-registering. + * @returns A map of registered tool names to their unregister disposers. + */ +export async function syncTools( + client: Client, + ctx: Context, + opts: ToolBridgeOptions, + previous: ToolDisposers, +): Promise { + for (const dispose of previous.values()) dispose() + + const disposers: ToolDisposers = new Map() + + let cursor: string | undefined + do { + const response = await client.listTools(cursor ? { cursor } : undefined) + for (const tool of response.tools) { + const registeredName = opts.toolPrefix + tool.name + const definition: ToolDefinition = { + name: registeredName, + description: tool.description ?? '', + parameters: tool.inputSchema, + execute: createExecutor(client, tool.name, opts), + } + try { + const dispose = ctx.tools.register(definition) + disposers.set(registeredName, dispose) + } catch { + // Name conflict — another tool with this name is already registered. + ctx.logger.warn(`mcp-client: skipping tool "${registeredName}" (name conflict)`) + } + } + cursor = response.nextCursor + } while (cursor) + + return disposers +} + +/** + * The shape we read from each MCP content block. Intentionally looser than the + * SDK's `ContentBlock` type: we're at a network trust boundary (data arrives + * from an external MCP server process via JSON-RPC), so fields that the SDK + * declares required may be absent at runtime if the server is buggy. + */ +interface McpContentBlock { + type: string + text?: string + mimeType?: string +} + +/** + * Create an execute function for one MCP tool. The executor calls + * `client.callTool` with abort signal and timeout, then maps the result + * to harness ContentBlocks. + * + * When the MCP server returns `isError: true`, the executor throws so that + * the ToolRegistry's catch path produces an `isError` result for the model. + */ +function createExecutor( + client: Client, + mcpToolName: string, + opts: ToolBridgeOptions, +): ToolDefinition['execute'] { + return async (args: unknown, exec: ToolExecution) => { + // The agent loop passes `JSON.parse(model_arguments)` which is usually an + // object, but can be any JSON value if the model misbehaves (outputs a bare + // string/number/null). Fallback to {} lets the MCP server produce a + // specific "missing required param" error the model can learn from. + const argsObj = (typeof args === 'object' && args !== null ? args : {}) as Record + const result = await client.callTool( + { name: mcpToolName, arguments: argsObj }, + undefined, + { + ...exec.signal ? { signal: exec.signal } : {}, + timeout: opts.toolCallTimeoutMs, + }, + ) + + // The SDK may return a legacy `toolResult` shape; normalize to content array. + if (!('content' in result) || !Array.isArray(result.content)) { + const text = 'toolResult' in result + ? JSON.stringify(result.toolResult) + : '(no output)' + return [{ type: 'text' as const, text }] + } + + // Trust boundary: the SDK's return type erases to `any[]` due to the + // union of CallToolResult | CompatibilityCallToolResult. We process each + // element defensively in extractText (reading only .type/.text/.mimeType + // with optional fallbacks). + // eslint-disable-next-line @typescript-eslint/no-unsafe-assignment + const content: McpContentBlock[] = result.content + const text = extractText(content, mcpToolName) + + // MCP isError → throw so ToolRegistry produces an isError result for the model. + if ('isError' in result && result.isError === true) { + throw new Error(text) + } + + return [{ type: 'text', text }] + } +} + +/** + * Extract text from an MCP content array into a single string. + * - text blocks: join with '\n' + * - image/audio/resource blocks: replaced with a placeholder + * + * Defensive: fields that the MCP spec declares required (mimeType, text) are + * guarded with fallbacks because this is a network trust boundary. + */ +function extractText(mcpContent: McpContentBlock[], toolName: string): string { + const parts: string[] = [] + + for (const block of mcpContent) { + switch (block.type) { + case 'text': + if (block.text !== undefined) parts.push(block.text) + break + case 'image': + parts.push(`[image: ${block.mimeType ?? 'unknown'}, content discarded]`) + break + case 'audio': + parts.push(`[audio: ${block.mimeType ?? 'unknown'}, content discarded]`) + break + case 'resource': + case 'resource_link': + parts.push('[resource: content discarded]') + break + default: + parts.push(`[unsupported content type: ${block.type}]`) + } + } + + return parts.join('\n') || `(${toolName} returned no text content)` +} diff --git a/packages/mcp/mcp-client/src/transport.ts b/packages/mcp/mcp-client/src/transport.ts new file mode 100644 index 0000000000..6f7c584b20 --- /dev/null +++ b/packages/mcp/mcp-client/src/transport.ts @@ -0,0 +1,56 @@ +/** + * Transport factory: creates the appropriate MCP transport based on the + * plugin's resolved config. Stdio spawns a child process (with credential + * scrubbing); Streamable HTTP connects to a URL. + * + * @module + */ + +import type { Transport } from '@modelcontextprotocol/sdk/shared/transport.js' +import { StdioClientTransport } from '@modelcontextprotocol/sdk/client/stdio.js' +import { StreamableHTTPClientTransport } from '@modelcontextprotocol/sdk/client/streamableHttp.js' +import type { Config } from './index.ts' + +/** + * Credential-shaped ambient env vars are NOT forwarded to the child by default + * (the parent harness's own secrets must not leak into a spawned process + * implicitly). Same pattern as `dsh-subagent-acp`. + */ +const SENSITIVE_ENV_PATTERN = /KEY|SECRET|TOKEN/i + +/** The ambient env minus credential-shaped vars, plus the spec's explicit env. */ +function buildChildEnv(extra: Record): Record { + const env: Record = {} + for (const [key, value] of Object.entries(process.env)) { + if (value !== undefined && !SENSITIVE_ENV_PATTERN.test(key)) env[key] = value + } + return { ...env, ...extra } +} + +/** + * Create an MCP transport from the resolved plugin config. + * + * @param config - Resolved plugin config discriminated on `transport`. + * @returns A connected-ready MCP Transport (stdio or Streamable HTTP). + */ +export function createTransport(config: Config): Transport { + switch (config.transport) { + case 'stdio': + return new StdioClientTransport({ + command: config.command, + args: config.args, + env: buildChildEnv(config.env), + cwd: config.cwd, + }) + case 'streamable-http': + // The MCP SDK's StreamableHTTPClientTransport has optional callback + // properties typed without `| undefined` (exactOptionalPropertyTypes + // mismatch with the Transport interface). The cast is safe — the SDK + // constructed the object, it simply doesn't declare the optionals + // strictly enough for our tsconfig. + return new StreamableHTTPClientTransport( + new URL(config.url), + { requestInit: { headers: config.headers } }, + ) as Transport + } +} diff --git a/packages/mcp/mcp-client/tests/apply.spec.ts b/packages/mcp/mcp-client/tests/apply.spec.ts new file mode 100644 index 0000000000..077cb4e3f6 --- /dev/null +++ b/packages/mcp/mcp-client/tests/apply.spec.ts @@ -0,0 +1,179 @@ +/** + * Tests for the mcp-client plugin's `apply` lifecycle entry point. + * Isolated file so vi.mock of the MCP SDK doesn't pollute other test suites. + */ +import { describe, expect, it, vi, beforeEach } from 'vitest' +import { Context } from 'cordis' +import SystemPrompt from '@deepseek-ai/dsh-system-prompt' +import ToolRegistry from '@deepseek-ai/dsh-tools' +import type { Config } from '@deepseek-ai/dsh-mcp-client' + +// ---- Mock MCP SDK ---- + +const mockConnect = vi.fn<() => Promise>() +const mockClose = vi.fn<() => Promise>() +const mockListTools = vi.fn() +const mockCallTool = vi.fn() +const mockSetNotificationHandler = vi.fn() + +class MockClient { + connect = mockConnect + close = mockClose + listTools = mockListTools + callTool = mockCallTool + setNotificationHandler = mockSetNotificationHandler +} + +vi.mock('@modelcontextprotocol/sdk/client/index.js', () => ({ + Client: MockClient, +})) + +vi.mock('@modelcontextprotocol/sdk/client/stdio.js', () => ({ + StdioClientTransport: vi.fn(), +})) + +vi.mock('@modelcontextprotocol/sdk/client/streamableHttp.js', () => ({ + StreamableHTTPClientTransport: vi.fn(), +})) + +// ---- Import under test (after mocks) ---- + +const { apply, name, inject, Config: ConfigSchema } = await import( + '@deepseek-ai/dsh-mcp-client/src/index.ts', +) + +// ---- Helpers ---- + +async function mountRegistry(): Promise { + const ctx = new Context() + await ctx.plugin(SystemPrompt) + await ctx.plugin(ToolRegistry) + return ctx +} + +const stdioConfig: Config = { + transport: 'stdio', + command: 'echo', + args: [], + env: {}, + cwd: '', + toolPrefix: '', + toolCallTimeoutMs: 60_000, +} + +// ---- Tests ---- + +describe('mcp-client plugin module exports', () => { + it('exports name, inject, and Config', () => { + expect(name).toBe('mcp-client') + expect(inject).toEqual(['tools']) + expect(ConfigSchema).toBeDefined() + }) +}) + +describe('apply (plugin lifecycle)', () => { + let ctx: Context + + beforeEach(async () => { + vi.clearAllMocks() + mockConnect.mockResolvedValue(undefined) + mockClose.mockResolvedValue(undefined) + mockListTools.mockResolvedValue({ + tools: [{ name: 'remote', description: 'A remote tool', inputSchema: { type: 'object' } }], + nextCursor: undefined, + }) + mockCallTool.mockResolvedValue({ content: [{ type: 'text', text: 'ok' }] }) + ctx = await mountRegistry() + }) + + it('connects, syncs tools, and registers a notification handler', async () => { + apply(ctx, stdioConfig) + await new Promise(r => setTimeout(r, 50)) + + expect(mockConnect).toHaveBeenCalled() + expect(mockListTools).toHaveBeenCalled() + expect(mockSetNotificationHandler).toHaveBeenCalled() + expect(ctx.tools.get('remote')).toBeDefined() + }) + + it('applies toolPrefix from config during sync', async () => { + apply(ctx, { ...stdioConfig, toolPrefix: 'mcp_' }) + await new Promise(r => setTimeout(r, 50)) + + expect(ctx.tools.get('mcp_remote')).toBeDefined() + expect(ctx.tools.get('remote')).toBeUndefined() + }) + + it('logs error and registers no tools when connect fails', async () => { + mockConnect.mockRejectedValue(new Error('connection refused')) + + apply(ctx, stdioConfig) + await new Promise(r => setTimeout(r, 50)) + + expect(mockListTools).not.toHaveBeenCalled() + expect(ctx.tools.get('remote')).toBeUndefined() + }) + + it('re-syncs tools on ToolListChanged notification', async () => { + apply(ctx, stdioConfig) + await new Promise(r => setTimeout(r, 50)) + + expect(ctx.tools.get('remote')).toBeDefined() + + // Simulate the notification handler being invoked with a new tool list. + mockListTools.mockResolvedValue({ + tools: [{ name: 'updated', inputSchema: { type: 'object' } }], + nextCursor: undefined, + }) + + // Extract and call the notification handler. + const handler = mockSetNotificationHandler.mock.calls[0]![1] as () => Promise + await handler() + + expect(ctx.tools.get('remote')).toBeUndefined() + expect(ctx.tools.get('updated')).toBeDefined() + }) + + it('effect disposer unregisters tools and closes client', async () => { + apply(ctx, stdioConfig) + await new Promise(r => setTimeout(r, 50)) + + expect(ctx.tools.get('remote')).toBeDefined() + + // Trigger disposal by disposing a child scope. + // Cordis ctx.effect registers the disposer; calling scope dispose runs it. + await ctx.fiber.dispose() + await new Promise(r => setTimeout(r, 50)) + + expect(mockClose).toHaveBeenCalled() + }) + + it('effect disposer handles client.close failure gracefully', async () => { + mockClose.mockRejectedValue(new Error('already closed')) + + apply(ctx, stdioConfig) + await new Promise(r => setTimeout(r, 50)) + + // Should not throw when dispose is triggered. + await ctx.fiber.dispose() + await new Promise(r => setTimeout(r, 50)) + + expect(mockClose).toHaveBeenCalled() + }) + + it('uses streamable-http config path', async () => { + const httpConfig: Config = { + transport: 'streamable-http', + url: 'http://localhost:3000/mcp', + headers: { Authorization: 'Bearer x' }, + toolPrefix: '', + toolCallTimeoutMs: 30_000, + } + + apply(ctx, httpConfig) + await new Promise(r => setTimeout(r, 50)) + + expect(mockConnect).toHaveBeenCalled() + expect(ctx.tools.get('remote')).toBeDefined() + }) +}) diff --git a/packages/mcp/mcp-client/tests/mcp-client.spec.ts b/packages/mcp/mcp-client/tests/mcp-client.spec.ts new file mode 100644 index 0000000000..e81369c0f0 --- /dev/null +++ b/packages/mcp/mcp-client/tests/mcp-client.spec.ts @@ -0,0 +1,518 @@ +import { describe, expect, it, vi, beforeEach } from 'vitest' +import { Context } from 'cordis' +import { CallId } from '@deepseek-ai/dsh-llm' +import SystemPrompt from '@deepseek-ai/dsh-system-prompt' +import ToolRegistry from '@deepseek-ai/dsh-tools' +import { syncTools, type ToolBridgeOptions } from '@deepseek-ai/dsh-mcp-client/src/tools.ts' +import { createTransport } from '@deepseek-ai/dsh-mcp-client/src/transport.ts' +import type { Config } from '@deepseek-ai/dsh-mcp-client' + +// ---- Mock MCP Client ---- + +interface MockTool { + name: string + description?: string + inputSchema: Record +} + +interface MockCallResult { + content: Array<{ type: string; text?: string; mimeType?: string }> + isError?: boolean +} + +function createMockClient(tools: MockTool[], callResult: MockCallResult = { content: [{ type: 'text', text: 'ok' }] }) { + return { + listTools: vi.fn().mockResolvedValue({ tools, nextCursor: undefined }), + callTool: vi.fn().mockResolvedValue(callResult), + setNotificationHandler: vi.fn(), + connect: vi.fn().mockResolvedValue(undefined), + close: vi.fn().mockResolvedValue(undefined), + } +} + +// ---- Test harness helper ---- + +async function mountRegistry(): Promise { + const ctx = new Context() + await ctx.plugin(SystemPrompt) + await ctx.plugin(ToolRegistry) + return ctx +} + +const defaultOpts: ToolBridgeOptions = { + toolPrefix: '', + toolCallTimeoutMs: 60_000, +} + +// ---- Tests ---- + +describe('syncTools', () => { + let ctx: Context + + beforeEach(async () => { + ctx = await mountRegistry() + }) + + it('registers tools from listTools response', async () => { + const client = createMockClient([ + { name: 'greet', description: 'Say hello', inputSchema: { type: 'object', properties: { name: { type: 'string' } } } }, + { name: 'add', description: 'Add numbers', inputSchema: { type: 'object', properties: {} } }, + ]) + + const disposers = await syncTools(client as never, ctx, defaultOpts, new Map()) + + expect(disposers.size).toBe(2) + expect(ctx.tools.get('greet')).toBeDefined() + expect(ctx.tools.get('add')).toBeDefined() + }) + + it('applies toolPrefix to registered names', async () => { + const client = createMockClient([ + { name: 'create_issue', description: 'Create an issue', inputSchema: { type: 'object' } }, + ]) + + const disposers = await syncTools(client as never, ctx, { ...defaultOpts, toolPrefix: 'gh_' }, new Map()) + + expect(disposers.size).toBe(1) + expect(ctx.tools.get('gh_create_issue')).toBeDefined() + expect(ctx.tools.get('create_issue')).toBeUndefined() + }) + + it('skips tools with conflicting names and logs warning', async () => { + // Pre-register a tool with the same name. + ctx.tools.register({ + name: 'existing', + description: 'Already here', + parameters: { type: 'object' }, + execute: async () => [{ type: 'text', text: 'native' }], + }) + + const client = createMockClient([ + { name: 'existing', description: 'Conflicts', inputSchema: { type: 'object' } }, + { name: 'unique', description: 'No conflict', inputSchema: { type: 'object' } }, + ]) + + const disposers = await syncTools(client as never, ctx, defaultOpts, new Map()) + + // Only the non-conflicting tool registers. + expect(disposers.size).toBe(1) + expect(ctx.tools.get('unique')).toBeDefined() + // Original tool unchanged. + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'existing', arguments: {} }) + expect(result.content[0]).toEqual({ type: 'text', text: 'native' }) + }) + + it('unregisters previous tools before re-syncing', async () => { + const client = createMockClient([ + { name: 'old_tool', inputSchema: { type: 'object' } }, + ]) + + const firstDisposers = await syncTools(client as never, ctx, defaultOpts, new Map()) + expect(ctx.tools.get('old_tool')).toBeDefined() + + // Second sync with different tools should remove old_tool. + client.listTools.mockResolvedValue({ tools: [{ name: 'new_tool', inputSchema: { type: 'object' } }], nextCursor: undefined }) + const secondDisposers = await syncTools(client as never, ctx, defaultOpts, firstDisposers) + + expect(ctx.tools.get('old_tool')).toBeUndefined() + expect(ctx.tools.get('new_tool')).toBeDefined() + expect(secondDisposers.size).toBe(1) + }) + + it('drains paginated listTools responses', async () => { + const client = createMockClient([]) + client.listTools + .mockResolvedValueOnce({ tools: [{ name: 'page1', inputSchema: { type: 'object' } }], nextCursor: 'cursor1' }) + .mockResolvedValueOnce({ tools: [{ name: 'page2', inputSchema: { type: 'object' } }], nextCursor: undefined }) + + const disposers = await syncTools(client as never, ctx, defaultOpts, new Map()) + + expect(disposers.size).toBe(2) + expect(ctx.tools.get('page1')).toBeDefined() + expect(ctx.tools.get('page2')).toBeDefined() + }) +}) + +describe('tool execution', () => { + let ctx: Context + + beforeEach(async () => { + ctx = await mountRegistry() + }) + + it('calls MCP callTool and returns text content', async () => { + const client = createMockClient( + [{ name: 'echo', inputSchema: { type: 'object' } }], + { content: [{ type: 'text', text: 'hello world' }] }, + ) + + await syncTools(client as never, ctx, defaultOpts, new Map()) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'echo', arguments: { msg: 'hi' } }) + + expect(result.isError).toBe(false) + expect(result.content).toEqual([{ type: 'text', text: 'hello world' }]) + expect(client.callTool).toHaveBeenCalledWith( + { name: 'echo', arguments: { msg: 'hi' } }, + undefined, + expect.objectContaining({ timeout: 60_000 }), + ) + }) + + it('joins multiple text blocks with newline', async () => { + const client = createMockClient( + [{ name: 'multi', inputSchema: { type: 'object' } }], + { content: [{ type: 'text', text: 'line1' }, { type: 'text', text: 'line2' }] }, + ) + + await syncTools(client as never, ctx, defaultOpts, new Map()) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'multi', arguments: {} }) + + expect(result.content).toEqual([{ type: 'text', text: 'line1\nline2' }]) + }) + + it('discards image content with placeholder', async () => { + const client = createMockClient( + [{ name: 'img', inputSchema: { type: 'object' } }], + { content: [{ type: 'text', text: 'before' }, { type: 'image', mimeType: 'image/png' }] }, + ) + + await syncTools(client as never, ctx, defaultOpts, new Map()) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'img', arguments: {} }) + + expect(result.content[0]).toEqual({ type: 'text', text: 'before\n[image: image/png, content discarded]' }) + }) + + it('maps isError to an error result via throw', async () => { + const client = createMockClient( + [{ name: 'fail', inputSchema: { type: 'object' } }], + { content: [{ type: 'text', text: 'something went wrong' }], isError: true }, + ) + + await syncTools(client as never, ctx, defaultOpts, new Map()) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'fail', arguments: {} }) + + expect(result.isError).toBe(true) + expect(result.content[0]).toEqual({ type: 'text', text: 'Error: something went wrong' }) + }) + + it('passes abort signal to callTool', async () => { + const controller = new AbortController() + const client = createMockClient( + [{ name: 'slow', inputSchema: { type: 'object' } }], + { content: [{ type: 'text', text: 'done' }] }, + ) + + await syncTools(client as never, ctx, defaultOpts, new Map()) + await ctx.tools.execute({ callId: CallId('c1'), name: 'slow', arguments: {}, signal: controller.signal }) + + expect(client.callTool).toHaveBeenCalledWith( + expect.anything(), + undefined, + expect.objectContaining({ signal: controller.signal }), + ) + }) + + it('handles legacy toolResult shape', async () => { + const client = createMockClient( + [{ name: 'legacy', inputSchema: { type: 'object' } }], + ) + client.callTool.mockResolvedValue({ toolResult: { key: 'value' } }) + + await syncTools(client as never, ctx, defaultOpts, new Map()) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'legacy', arguments: {} }) + + expect(result.isError).toBe(false) + expect(result.content[0]).toEqual({ type: 'text', text: '{"key":"value"}' }) + }) +}) + +describe('tool execution edge cases', () => { + let ctx: Context + + beforeEach(async () => { + ctx = await mountRegistry() + }) + + it('handles audio content with placeholder', async () => { + const client = createMockClient( + [{ name: 'audio_tool', inputSchema: { type: 'object' } }], + { content: [{ type: 'audio', mimeType: 'audio/mp3' }] }, + ) + + await syncTools(client as never, ctx, defaultOpts, new Map()) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'audio_tool', arguments: {} }) + + expect(result.content[0]).toEqual({ type: 'text', text: '[audio: audio/mp3, content discarded]' }) + }) + + it('handles resource content with placeholder', async () => { + const client = createMockClient( + [{ name: 'res_tool', inputSchema: { type: 'object' } }], + { content: [{ type: 'resource' }] }, + ) + + await syncTools(client as never, ctx, defaultOpts, new Map()) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'res_tool', arguments: {} }) + + expect(result.content[0]).toEqual({ type: 'text', text: '[resource: content discarded]' }) + }) + + it('handles resource_link content with placeholder', async () => { + const client = createMockClient( + [{ name: 'link_tool', inputSchema: { type: 'object' } }], + { content: [{ type: 'resource_link' }] }, + ) + + await syncTools(client as never, ctx, defaultOpts, new Map()) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'link_tool', arguments: {} }) + + expect(result.content[0]).toEqual({ type: 'text', text: '[resource: content discarded]' }) + }) + + it('handles unknown content types', async () => { + const client = createMockClient( + [{ name: 'unknown_tool', inputSchema: { type: 'object' } }], + { content: [{ type: 'video' }] }, + ) + + await syncTools(client as never, ctx, defaultOpts, new Map()) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'unknown_tool', arguments: {} }) + + expect(result.content[0]).toEqual({ type: 'text', text: '[unsupported content type: video]' }) + }) + + it('handles image with missing mimeType (buggy server)', async () => { + const client = createMockClient( + [{ name: 'img2', inputSchema: { type: 'object' } }], + { content: [{ type: 'image' }] }, + ) + + await syncTools(client as never, ctx, defaultOpts, new Map()) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'img2', arguments: {} }) + + expect(result.content[0]).toEqual({ type: 'text', text: '[image: unknown, content discarded]' }) + }) + + it('handles audio with missing mimeType (buggy server)', async () => { + const client = createMockClient( + [{ name: 'audio_no_mime', inputSchema: { type: 'object' } }], + { content: [{ type: 'audio' }] }, + ) + + await syncTools(client as never, ctx, defaultOpts, new Map()) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'audio_no_mime', arguments: {} }) + + expect(result.content[0]).toEqual({ type: 'text', text: '[audio: unknown, content discarded]' }) + }) + + it('handles text block with missing text (buggy server)', async () => { + const client = createMockClient( + [{ name: 'notext', inputSchema: { type: 'object' } }], + { content: [{ type: 'text' }] }, + ) + + await syncTools(client as never, ctx, defaultOpts, new Map()) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'notext', arguments: {} }) + + expect(result.content[0]).toEqual({ type: 'text', text: '(notext returned no text content)' }) + }) + + it('handles empty content array', async () => { + const client = createMockClient( + [{ name: 'empty_tool', inputSchema: { type: 'object' } }], + { content: [] }, + ) + + await syncTools(client as never, ctx, defaultOpts, new Map()) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'empty_tool', arguments: {} }) + + expect(result.content[0]).toEqual({ type: 'text', text: '(empty_tool returned no text content)' }) + }) + + + it('handles legacy toolResult with undefined value', async () => { + const client = createMockClient( + [{ name: 'legacy2', inputSchema: { type: 'object' } }], + ) + client.callTool.mockResolvedValue({}) + + await syncTools(client as never, ctx, defaultOpts, new Map()) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'legacy2', arguments: {} }) + + expect(result.content[0]).toEqual({ type: 'text', text: '(no output)' }) + }) + + it('handles isError with non-text content (fallback error message)', async () => { + const client = createMockClient( + [{ name: 'err_notext', inputSchema: { type: 'object' } }], + { content: [{ type: 'image', mimeType: 'image/png' }], isError: true }, + ) + + await syncTools(client as never, ctx, defaultOpts, new Map()) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'err_notext', arguments: {} }) + + expect(result.isError).toBe(true) + // The error message falls back to 'MCP tool error' when content[0] is not text. + // But mapContent converts image to text placeholder, so it should use that. + // Actually mapContent ALWAYS returns text, so the ternary always takes the truthy branch. + // Let me check: mapContent returns [{type:'text', text:'[image: ...]'}], so content[0].type IS 'text'. + expect(result.content[0]).toEqual({ type: 'text', text: 'Error: [image: image/png, content discarded]' }) + }) + + + it('uses tool description when provided', async () => { + const client = createMockClient([ + { name: 'described', description: 'A described tool', inputSchema: { type: 'object' } }, + ]) + + await syncTools(client as never, ctx, defaultOpts, new Map()) + const tool = ctx.tools.get('described') + expect(tool?.description).toBe('A described tool') + }) + + it('uses empty description when tool has no description', async () => { + const client = createMockClient([ + { name: 'nodesc', inputSchema: { type: 'object' } }, + ]) + + await syncTools(client as never, ctx, defaultOpts, new Map()) + const tool = ctx.tools.get('nodesc') + expect(tool?.description).toBe('') + }) +}) + +describe('createTransport', () => { + it('creates StdioClientTransport for stdio config', () => { + const config: Config = { + transport: 'stdio', + command: 'node', + args: ['server.js'], + env: {}, + cwd: '/tmp', + toolPrefix: '', + toolCallTimeoutMs: 60_000, + } + const transport = createTransport(config) + expect(transport).toBeDefined() + expect(transport).toHaveProperty('start') + expect(transport).toHaveProperty('close') + }) + + it('creates StreamableHTTPClientTransport for http config without headers', () => { + const config: Config = { + transport: 'streamable-http', + url: 'http://localhost:3000/mcp', + headers: {}, + toolPrefix: '', + toolCallTimeoutMs: 60_000, + } + const transport = createTransport(config) + expect(transport).toBeDefined() + expect(transport).toHaveProperty('start') + expect(transport).toHaveProperty('close') + }) + + it('creates StreamableHTTPClientTransport for http config with headers', () => { + const config: Config = { + transport: 'streamable-http', + url: 'http://localhost:3000/mcp', + headers: { Authorization: 'Bearer token' }, + toolPrefix: '', + toolCallTimeoutMs: 60_000, + } + const transport = createTransport(config) + expect(transport).toBeDefined() + expect(transport).toHaveProperty('start') + expect(transport).toHaveProperty('close') + }) + + it('scrubs sensitive env vars and forwards the rest', () => { + const original = { ...process.env } + try { + process.env.SAFE_VAR = 'kept' + process.env.MY_SECRET = 'hidden' + process.env.API_KEY = 'hidden' + process.env.AUTH_TOKEN = 'hidden' + + const config: Config = { + transport: 'stdio', + command: 'echo', + args: [], + env: { EXTRA: 'injected' }, + cwd: '', + toolPrefix: '', + toolCallTimeoutMs: 60_000, + } + // createTransport internally calls buildChildEnv; we verify by inspecting + // the constructed StdioClientTransport. Since we can't inspect private fields + // easily, we at least confirm it doesn't throw and returns a transport. + const transport = createTransport(config) + expect(transport).toBeDefined() + } finally { + // Restore env + delete process.env.SAFE_VAR + delete process.env.MY_SECRET + delete process.env.API_KEY + delete process.env.AUTH_TOKEN + for (const key of Object.keys(process.env)) { + if (!(key in original)) Reflect.deleteProperty(process.env, key) + } + } + }) + + it('merges explicit env on top of scrubbed ambient env', () => { + const config: Config = { + transport: 'stdio', + command: 'echo', + args: [], + env: { CUSTOM: 'value' }, + cwd: '', + toolPrefix: '', + toolCallTimeoutMs: 60_000, + } + const transport = createTransport(config) + expect(transport).toBeDefined() + }) +}) + +describe('tool execution — non-object args fallback', () => { + let ctx: Context + + beforeEach(async () => { + ctx = await mountRegistry() + }) + + it('coerces null args to empty object for callTool', async () => { + const client = createMockClient( + [{ name: 'coerce', inputSchema: { type: 'object' } }], + { content: [{ type: 'text', text: 'ok' }] }, + ) + + await syncTools(client as never, ctx, defaultOpts, new Map()) + // Simulate model emitting `null` as tool arguments (malformed). + await ctx.tools.execute({ callId: CallId('c1'), name: 'coerce', arguments: null }) + + expect(client.callTool).toHaveBeenCalledWith( + { name: 'coerce', arguments: {} }, + undefined, + expect.anything(), + ) + }) + + it('coerces primitive string args to empty object for callTool', async () => { + const client = createMockClient( + [{ name: 'coerce2', inputSchema: { type: 'object' } }], + { content: [{ type: 'text', text: 'ok' }] }, + ) + + await syncTools(client as never, ctx, defaultOpts, new Map()) + await ctx.tools.execute({ callId: CallId('c1'), name: 'coerce2', arguments: 'bad' }) + + expect(client.callTool).toHaveBeenCalledWith( + { name: 'coerce2', arguments: {} }, + undefined, + expect.anything(), + ) + }) +}) + diff --git a/packages/mcp/mcp-client/tsconfig.json b/packages/mcp/mcp-client/tsconfig.json new file mode 100644 index 0000000000..e9c9266415 --- /dev/null +++ b/packages/mcp/mcp-client/tsconfig.json @@ -0,0 +1,15 @@ +{ + "extends": "../../../tsconfig.base.json", + "compilerOptions": { + "rootDir": "src", + "outDir": "lib/types" + }, + "include": ["src"], + "references": [ + { "path": "../../../vendor/cosmokit" }, + { "path": "../../../vendor/cordis" }, + { "path": "../../../vendor/schemastery" }, + { "path": "../../llm/llm" }, + { "path": "../../core/tools" } + ] +} diff --git a/pnpm-lock.yaml b/pnpm-lock.yaml index 8f6c140ee7..a1752a60f2 100644 --- a/pnpm-lock.yaml +++ b/pnpm-lock.yaml @@ -496,7 +496,7 @@ importers: dependencies: '@earendil-works/pi-ai': specifier: ^0.79.1 - version: 0.79.3(ws@8.21.0)(zod@4.4.3) + version: 0.79.3(@modelcontextprotocol/sdk@1.29.0(zod@4.4.3))(ws@8.21.0)(zod@4.4.3) schemastery: specifier: ^3.18.0 version: 3.18.0 @@ -511,6 +511,25 @@ importers: specifier: ^4.0.0-rc.6 version: 4.0.0-rc.6(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.4) + packages/mcp/mcp-client: + dependencies: + '@modelcontextprotocol/sdk': + specifier: ^1.12.0 + version: 1.29.0(zod@4.4.3) + schemastery: + specifier: ^3.18.0 + version: 3.18.0 + devDependencies: + '@deepseek-ai/dsh-llm': + specifier: workspace:^ + version: link:../../llm/llm + '@deepseek-ai/dsh-tools': + specifier: workspace:^ + version: link:../../core/tools + cordis: + specifier: ^4.0.0-rc.6 + version: 4.0.0-rc.6(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.4) + packages/session-persistence/session-persistence: devDependencies: '@deepseek-ai/dsh-session': @@ -1656,6 +1675,12 @@ packages: '@modelcontextprotocol/sdk': optional: true + '@hono/node-server@1.19.14': + resolution: {integrity: sha512-GwtvgtXxnWsucXvbQXkRgqksiH2Qed37H9xHZocE5sA3N8O8O8/8FA3uclQXxXVzc9XBZuEOMK7+r02FmSpHtw==} + engines: {node: '>=18.14.1'} + peerDependencies: + hono: ^4 + '@humanfs/core@0.19.2': resolution: {integrity: sha512-UhXNm+CFMWcbChXywFwkmhqjs3PRCmcSa/hfBgLIb7oQ5HNb1wS0icWsGtSAUNgefHeI+eBrA8I1fxmbHsGdvA==} engines: {node: '>=18.18.0'} @@ -1701,6 +1726,16 @@ packages: '@mistralai/mistralai@2.2.1': resolution: {integrity: sha512-uKU8CZmL2RzYKmplsU01hii4p3pe4HqJefpWNRWXm1Tcm0Sm4xXfwSLIy4k7ZCPlbETCGcp69E7hZs+WOJ5itQ==} + '@modelcontextprotocol/sdk@1.29.0': + resolution: {integrity: sha512-zo37mZA9hJWpULgkRpowewez1y6ML5GsXJPY8FI0tBBCd77HEvza4jDqRKOXgHNn867PVGCyTdzqpz0izu5ZjQ==} + engines: {node: '>=18'} + peerDependencies: + '@cfworker/json-schema': ^4.1.1 + zod: ^3.25 || ^4.0 + peerDependenciesMeta: + '@cfworker/json-schema': + optional: true + '@napi-rs/wasm-runtime@1.1.5': resolution: {integrity: sha512-AWPoBRJ9tsnVhor4sjO7rkni+7p+2IAEFj6cx06UgP10jkQHqay/36uRV/bFkgrh18D9vb4cr8Q0Pthskgzy+Q==} peerDependencies: @@ -2472,6 +2507,10 @@ packages: '@vitest/utils@4.1.8': resolution: {integrity: sha512-uOJamYALNhfJ6iolExyQM40yIQwDqYnkKtQ5VCiSe17E33H0aQ/u+1GlRuz4LZBk6Mm3sg90G9hEbmEt37C1Zg==} + accepts@2.0.0: + resolution: {integrity: sha512-5cvg6CtKwfgdmVqY1WIiXKc3Q1bkRqGLi+2W/6ao+6Y7gu/RCwRuAhGEzh5B4KlszSuTLgZYuqFqo5bImjNKng==} + engines: {node: '>= 0.6'} + acorn-jsx@5.3.2: resolution: {integrity: sha512-rq9s+JNhf0IChjtDXxllJ7g41oZk5SlXtp0LHwyA5cejwn7vKmKp4pPri6YEePv2PU65sAsegbXtIinmDFDXgQ==} peerDependencies: @@ -2486,9 +2525,20 @@ packages: resolution: {integrity: sha512-MnA+YT8fwfJPgBx3m60MNqakm30XOkyIoH1y6huTQvC0PwZG7ki8NacLBcrPbNoo8vEZy7Jpuk7+jMO+CUovTQ==} engines: {node: '>= 14'} + ajv-formats@3.0.1: + resolution: {integrity: sha512-8iUql50EUR+uUcdRQ3HDqa6EVyo3docL8g5WJ3FNcWmu62IbkGUue/pEyLBW8VGKKucTPgqeks4fIU1DA4yowQ==} + peerDependencies: + ajv: ^8.0.0 + peerDependenciesMeta: + ajv: + optional: true + ajv@6.15.0: resolution: {integrity: sha512-fgFx7Hfoq60ytK2c7DhnF8jIvzYgOMxfugjLOSMHjLIPgenqa7S7oaagATUq99mV6IYvN2tRmC0wnTYX6iPbMw==} + ajv@8.20.0: + resolution: {integrity: sha512-Thbli+OlOj+iMPYFBVBfJ3OmCAnaSyNn4M1vz9T6Gka5Jt9ba/HIR56joy65tY6kx/FCF5VXNB819Y7/GUrBGA==} + ansis@4.3.1: resolution: {integrity: sha512-BJ8/l4R5LRE7hW9WdSuGYrLSHi2ynxeFpDFbH0K/CgNeY/tyhk+vO6TYxXC5r5CpUhNVX310xzPsN/H9lCdfOA==} engines: {node: '>=14'} @@ -2526,6 +2576,10 @@ packages: birpc@4.0.0: resolution: {integrity: sha512-LShSxJP0KTmd101b6DRyGBj57LZxSDYWKitQNW/mi8GRMvZb078Uf9+pveax1DrVL89vm7mWe+TovdI/UDOuPw==} + body-parser@2.3.0: + resolution: {integrity: sha512-2cGmJupaNgg+QUwVLAucDuWuoMZ6EX9iHDRswZ5lsNYEmwPaRknMPCLZz07yTzVq/83p4o/wzbDZbBrTvGGTIw==} + engines: {node: '>=18'} + bowser@2.14.1: resolution: {integrity: sha512-tzPjzCxygAKWFOJP011oxFHs57HzIhOEracIgAePE4pqB3LikALKnSzUyU4MGs9/iCEUuHlAJTjTc5M+u7YEGg==} @@ -2536,10 +2590,22 @@ packages: buffer-equal-constant-time@1.0.1: resolution: {integrity: sha512-zRpUiDwd/xk6ADqPMATG8vc9VPrkck7T07OIx0gnjmJAnHnTVXNQG3vfvWNuiZIkwu9KrKdA1iJKfsfTVxE6NA==} + bytes@3.1.2: + resolution: {integrity: sha512-/Nf7TyzTx6S3yRJObOAV7956r8cr2+Oj8AC5dt8wSP3BQAoeX58NoHyCU8P8zGkNXStjTSi6fzO6F0pBdcYbEg==} + engines: {node: '>= 0.8'} + cac@7.0.0: resolution: {integrity: sha512-tixWYgm5ZoOD+3g6UTea91eow5z6AAHaho3g0V9CNSNb45gM8SmflpAc+GRd1InC4AqN/07Unrgp56Y94N9hJQ==} engines: {node: '>=20.19.0'} + call-bind-apply-helpers@1.0.2: + resolution: {integrity: sha512-Sp1ablJ0ivDkSzjcaJdxEunN5/XvksFJ2sMBFfq6x0ryhQV/2b/KwFe21cMpmHtPOSij8K99/wSfoEuTObmuMQ==} + engines: {node: '>= 0.4'} + + call-bound@1.0.4: + resolution: {integrity: sha512-+ys997U96po4Kx/ABpBCqhA9EuxJaQWDQg7295H4hBphv3IZg0boBKuwYpt4YXp6MZ5AmZQnU/tyMTlRpaSejg==} + engines: {node: '>= 0.4'} + ccount@2.0.1: resolution: {integrity: sha512-eyrF0jiFpY+3drT6383f1qhkbGsLSifNAjA61IUjZjmLCWjItY6LB9ft9YhoDgwfmclB2zhu51Lc7+95b8NRAg==} @@ -2562,9 +2628,29 @@ packages: resolution: {integrity: sha512-OkTL9umf+He2DZkUq8f8J9of7yL6RJKI24dVITBmNfZBmri9zYZQrKkuXiKhyfPSu8tUhnVBB1iKXevvnlR4Ww==} engines: {node: '>= 12'} + content-disposition@1.1.0: + resolution: {integrity: sha512-5jRCH9Z/+DRP7rkvY83B+yGIGX96OYdJmzngqnw2SBSxqCFPd0w2km3s5iawpGX8krnwSGmF0FW5Nhr0Hfai3g==} + engines: {node: '>=18'} + + content-type@1.0.5: + resolution: {integrity: sha512-nTjqfcBFEipKdXCv4YDQWCfmcLZKm81ldF0pAopTvyrFGVbcR6P/VAAd5G7N+0tTr8QqiU0tFadD6FK4NtJwOA==} + engines: {node: '>= 0.6'} + + content-type@2.0.0: + resolution: {integrity: sha512-j/O/d7GcZCyNl7/hwZAb606rzqkyvaDctLmckbxLzHvFBzTJHuGEdodATcP3yIRoDrLHkIATJuvzbFlp/ki2cQ==} + engines: {node: '>=18'} + convert-source-map@2.0.0: resolution: {integrity: sha512-Kvp459HrV2FEJ1CAsi1Ku+MY3kasH19TFykTz2xWmMeq6bk2NU3XXvfJ+Q61m0xktWwt+1HSYf3JZsTms3aRJg==} + cookie-signature@1.2.2: + resolution: {integrity: sha512-D76uU73ulSXrD1UXF4KE2TMxVVwhsnCgfAyTg9k8P6KGZjlXKrOLe4dJQKI3Bxi5wjesZoFXJWElNWBjPZMbhg==} + engines: {node: '>=6.6.0'} + + cookie@0.7.2: + resolution: {integrity: sha512-yki5XnKuf750l50uGTllt6kKILY4nQ1eNIQatoXEByZ5dWgnKqbnqmTrBE5B4N7lrMJKQ2ytWMiTO2o0v6Ew/w==} + engines: {node: '>= 0.6'} + cordis@4.0.0-rc.6: resolution: {integrity: sha512-GzUv7zCKh3FlgM3/Ad2S03UpYO3v4u1GcKa7ig4K2je4lCrgJ/S64ziiZI6XNyKEa1tZwdzj4oBQrhYDLgfEiA==} hasBin: true @@ -2577,6 +2663,10 @@ packages: '@cordisjs/plugin-loader': optional: true + cors@2.8.6: + resolution: {integrity: sha512-tJtZBBHA6vjIAaF6EnIaq6laBBP9aq/Y3ouVJjEfoHbRBcHBAHYcMh/w8LDrk2PvIMMq8gmopa5D4V8RmbrxGw==} + engines: {node: '>= 0.10'} + cose-base@1.0.3: resolution: {integrity: sha512-s9whTXInMSgAp/NVXVNuVxVKzGH2qck3aQlVHxDCdAEPgtMKwc4Wq6/QKhgdEdgbLSi9rBTAcPoRa6JpiG4ksg==} @@ -2785,6 +2875,10 @@ packages: delaunator@5.1.0: resolution: {integrity: sha512-AGrQ4QSgssa1NGmWmLPqN5NY2KajF5MqxetNEO+o0n3ZwZZeTmt7bBnvzHWrmkZFxGgr4HdyFgelzgi06otLuQ==} + depd@2.0.0: + resolution: {integrity: sha512-g7nH6P6dyDioJogAAGprGpCtVImJhpPk/roCzdb3fIh61/s/nPsfR6onyMwkCAR/OlC3yBC0lESvUoQEAssIrw==} + engines: {node: '>= 0.8'} + dequal@2.0.3: resolution: {integrity: sha512-0je+qPKHEMohvfRTCEo3CrPG6cAzAYgmzKyxRiYSSDkS6eGJdyVJm7WaYA5ECaAD9wLB2T4EEeymA5aFVcYXCA==} engines: {node: '>=6'} @@ -2812,20 +2906,43 @@ packages: oxc-resolver: optional: true + dunder-proto@1.0.1: + resolution: {integrity: sha512-KIN/nDJBQRcXw0MLVhZE9iQHmG68qAVIBg9CqmUYjmQIhgij9U5MFvrqkUL5FbtyyzZuOeOt0zdeRe4UY7ct+A==} + engines: {node: '>= 0.4'} + ecdsa-sig-formatter@1.0.11: resolution: {integrity: sha512-nagl3RYrbNv6kQkeJIpt6NJZy8twLB/2vtz6yN9Z4vRKHN4/QZJIEbqohALSgwKdnksuY3k5Addp5lg8sVoVcQ==} + ee-first@1.1.1: + resolution: {integrity: sha512-WMwm9LhRUo+WUaRN+vRuETqG89IgZphVSNkdFgeb6sS/E4OrDIN7t48CAewSHXc6C8lefD8KKfr5vY61brQlow==} + empathic@2.0.1: resolution: {integrity: sha512-YGRs8knHhKHVShLkFET/rWAU8kmHbOV5LwN938RHI0pljAJ1Gf6SzXsSmRaEzcXTtOOmVqJ5+WtQPL5uigY50Q==} engines: {node: '>=14'} + encodeurl@2.0.0: + resolution: {integrity: sha512-Q0n9HRi4m6JuGIV1eFlmvJB7ZEVxu93IrMyiMsGC0lrMJMWzRgx6WGquyfQgZVb31vhGgXnfmPNNXmxnOkRBrg==} + engines: {node: '>= 0.8'} + entities@8.0.0: resolution: {integrity: sha512-zwfzJecQ/Uej6tusMqwAqU/6KL2XaB2VZ2Jg54Je6ahNBGNH6Ek6g3jjNCF0fG9EWQKGZNddNjU5F1ZQn/sBnA==} engines: {node: '>=20.19.0'} + es-define-property@1.0.1: + resolution: {integrity: sha512-e3nRfgfUZ4rNGL232gUgX06QNyyez04KdjFrF+LTRoOXmrOgFKDg4BCdsjW8EnT69eqdYGmRpJwiPVYNrCaW3g==} + engines: {node: '>= 0.4'} + + es-errors@1.3.0: + resolution: {integrity: sha512-Zf5H2Kxt2xjTvbJvP2ZWLEICxA6j+hAmMzIlypy4xcBg1vKVnx89Wy0GbS+kf5cwCVFFzdCFh2XSCFNULS6csw==} + engines: {node: '>= 0.4'} + es-module-lexer@2.1.0: resolution: {integrity: sha512-n27zTYMjYu1aj4MjCWzSP7G9r75utsaoc8m61weK+W8JMBGGQybd43GstCXZ3WNmSFtGT9wi59qQTW6mhTR5LQ==} + es-object-atoms@1.1.2: + resolution: {integrity: sha512-HWcBoN6NileqtSydK2FqHbS/LoDd2pqrnQHLyJzBj4kOp/ky2MWMN694xOfkK8/SnUsW2DH7EfyVlydKCsm1Zw==} + engines: {node: '>= 0.4'} + es-toolkit@1.49.0: resolution: {integrity: sha512-G5iZ6Pc/FNRY/soKZHC+TxGDD83rHUDXxzaWhGCX44vAv/tMs56WMusnm/KMNK+luUPsgA9U28cGr4RDlSzL2g==} @@ -2834,6 +2951,9 @@ packages: engines: {node: '>=18'} hasBin: true + escape-html@1.0.3: + resolution: {integrity: sha512-NiSupZ4OeuGwr68lGIeym/ksIZMJodUGOSCZ/FSnTxcrekbvqrgdUxlJOMpijaKZVjAJrWrGs/6Jy8OMuyj9ow==} + escape-string-regexp@4.0.0: resolution: {integrity: sha512-TtpcNJ3XAzx3Gq8sWRzJaVajRs0uVxA2YAkdb1jm2YkPz4G6egUFAyA3n5vtEIZefPk5Wa4UXbKuS5fKkJWdgA==} engines: {node: '>=10'} @@ -2895,10 +3015,32 @@ packages: resolution: {integrity: sha512-kVscqXk4OCp68SZ0dkgEKVi6/8ij300KBWTJq32P/dYeWTSwK41WyTxalN1eRmA5Z9UU/LX9D7FWSmV9SAYx6g==} engines: {node: '>=0.10.0'} + etag@1.8.1: + resolution: {integrity: sha512-aIL5Fx7mawVa300al2BnEE4iNvo1qETxLrPI/o05L7z6go7fCw1J6EQmbK4FmJ2AS7kgVF/KEZWufBfdClMcPg==} + engines: {node: '>= 0.6'} + + eventsource-parser@3.1.0: + resolution: {integrity: sha512-kJezFj9YFAMLeORyi7aCLxLbD5/qWMQnoMVlVPyHIll7lgRJCc3JVln9Vgl9nwQi0YkMnhdGTMNn7CkRRAptMg==} + engines: {node: '>=18.0.0'} + + eventsource@3.0.7: + resolution: {integrity: sha512-CRT1WTyuQoD771GW56XEZFQ/ZoSfWid1alKGDYMmkt2yl8UXrVR4pspqWNEcqKvVIzg6PAltWjxcSSPrboA4iA==} + engines: {node: '>=18.0.0'} + expect-type@1.3.0: resolution: {integrity: sha512-knvyeauYhqjOYvQ66MznSMs83wmHrCycNEN6Ao+2AeYEfxUIkuiVxdEa1qlGEPK+We3n0THiDciYSsCcgW/DoA==} engines: {node: '>=12.0.0'} + express-rate-limit@8.5.2: + resolution: {integrity: sha512-5Kb34ipNX694DH48vN9irak1Qx30nb0PLYHXfJgw4YEjiC3ZEmZJhwOp+VfiCYwFzvFTdB9QkArYS5kXa2cx2A==} + engines: {node: '>= 16'} + peerDependencies: + express: '>= 4.11' + + express@5.2.1: + resolution: {integrity: sha512-hIS4idWWai69NezIdRt2xFVofaF4j+6INOpJlVOLDO8zXGpUVEVzIYk12UUi2JzjEzWL3IOAxcTubgz9Po0yXw==} + engines: {node: '>= 18'} + extend@3.0.2: resolution: {integrity: sha512-fjquC59cD7CyW6urNXK0FBufkZcoiGG80wTuPujX590cB5Ttln20E2UB4S/WARVqhXffZl2LNgS+gQdPIIim/g==} @@ -2915,6 +3057,9 @@ packages: fast-levenshtein@2.0.6: resolution: {integrity: sha512-DCXu6Ifhqcks7TZKY3Hxp3y6qphY5SJZmrWMDrKcERSOXWQdMhU9Ig/PYrzyw/ul9jOIyh0N4M0tbC5hodg8dw==} + fast-uri@3.1.3: + resolution: {integrity: sha512-i70LwGWUduXqzicKXWshooq+sWL1K3WUU5rKZNG/0i3a1OSoX3HqhH5WbWwTmqWfor4urUakGPiRQcleRZTwOg==} + fast-xml-builder@1.2.0: resolution: {integrity: sha512-00aAWieqff+ZJhsXA4g1g7M8k+7AYoMUUHF+/zFb5U6Uv/P0Vl4QZo84/IcufzYalLuEj9928bXN9PbbFzMF0Q==} @@ -2942,6 +3087,10 @@ packages: resolution: {integrity: sha512-XXTUwCvisa5oacNGRP9SfNtYBNAMi+RPwBFmblZEF7N7swHYQS6/Zfk7SRwx4D5j3CH211YNRco1DEMNVfZCnQ==} engines: {node: '>=16.0.0'} + finalhandler@2.1.1: + resolution: {integrity: sha512-S8KoZgRZN+a5rNwqTxlZZePjT/4cnm0ROV70LedRHZ0p8u9fRID0hJUZQpkKLzro8LfmC8sx23bY6tVNxv8pQA==} + engines: {node: '>= 18.0.0'} + find-up@5.0.0: resolution: {integrity: sha512-78/PXT1wlLLDgTzDs7sjq9hzz0vXD+zn+7wypEe4fXQxCmdmqfGsEPQxmiCSQI3ajFV91bVSsvNtrJRiW6nGng==} engines: {node: '>=10'} @@ -2962,11 +3111,22 @@ packages: resolution: {integrity: sha512-buewHzMvYL29jdeQTVILecSaZKnt/RJWjoZCF5OW60Z67/GmSLBkOFM7qh1PI3zFNtJbaZL5eQu1vLfazOwj4g==} engines: {node: '>=12.20.0'} + forwarded@0.2.0: + resolution: {integrity: sha512-buRG0fpBtRHSTCOASe6hD258tEubFoRLb4ZNA6NxMVHNw2gOcwHo9wyablzMzOA5z9xA9L1KNjk/Nt6MT9aYow==} + engines: {node: '>= 0.6'} + + fresh@2.0.0: + resolution: {integrity: sha512-Rx/WycZ60HOaqLKAi6cHRKKI7zxWbJ31MhntmtwMoaTeF7XFH9hhBp8vITaMidfljRQ6eYWCKkaTK+ykVJHP2A==} + engines: {node: '>= 0.8'} + fsevents@2.3.3: resolution: {integrity: sha512-5xoDfX+fL7faATnagmWPpbFtwh/R77WmMMqqHGS65C3vvB0YHrgF+B1YmZ3441tMj5n63k0212XNoJwzlhffQw==} engines: {node: ^8.16.0 || ^10.6.0 || >=11.0.0} os: [darwin] + function-bind@1.1.2: + resolution: {integrity: sha512-7XHNxH7qX9xG5mIwxkhumTox/MIRNcOgDrxWsMt2pAr23WHp6MrRlN7FBSFpCpr+oVO0F744iUgR82nJMfG2SA==} + gaxios@7.1.5: resolution: {integrity: sha512-5FZy72Rh8LhtjmvDrKkI+lVhrsQrVKVsItxMoDm5mNQE+xR0WVIIs+jzPSJgBvKVsLi24fZhXJIsNI0bihDzFg==} engines: {node: '>=18'} @@ -2975,6 +3135,14 @@ packages: resolution: {integrity: sha512-zV/5HKTfCeKWnxG0Dmrw51hEWFGfcF2xiXqcA3+J90WDuP0SvoiSO5ORvcBsifmx/FoIjgQN3oNOGaQ5PhLFkg==} engines: {node: '>=18'} + get-intrinsic@1.3.0: + resolution: {integrity: sha512-9fSjSaos/fRIVIp+xSJlE6lfwhES7LNtKaCBIamHsjr2na1BiABJPo0mOjjz8GJDURarmCPGqaiVg5mfjb98CQ==} + engines: {node: '>= 0.4'} + + get-proto@1.0.1: + resolution: {integrity: sha512-sTSfBjoXBp89JvIKIefqw7U2CCebsc74kiY6awiGogKtoSGbgjYE/G/+l9sF3MWFPNc9IcoOC4ODfKHfxFmp0g==} + engines: {node: '>= 0.4'} + get-tsconfig@4.14.0: resolution: {integrity: sha512-yTb+8DXzDREzgvYmh6s9vHsSVCHeC0G3PI5bEXNBHtmshPnO+S5O7qgLEOn0I5QvMy6kpZN8K1NKGyilLb93wA==} @@ -2997,6 +3165,10 @@ packages: resolution: {integrity: sha512-eAmLkjDjAFCVXg7A1unxHsLf961m6y17QFqXqAXGj/gVkKFrEICfStRfwUlGNfeCEjNRa32JEWOUTlYXPyyKvA==} engines: {node: '>=14'} + gopd@1.2.0: + resolution: {integrity: sha512-ZUKRh6/kUFoAiTAtTYPZJ3hw9wNxx+BIBOijnlG9PnrJsCcSjs1wyyD6vJpaYtgnzDrKYRSqf3OO6Rfa93xsRg==} + engines: {node: '>= 0.4'} + hachure-fill@0.5.2: resolution: {integrity: sha512-3GKBOn+m2LX9iq+JC1064cSFprJY4jL1jCXTcpnfER5HYE2l/4EfWSGzkPa/ZDBmYI0ZOEj5VHV/eKnPGkHuOg==} @@ -3004,6 +3176,18 @@ packages: resolution: {integrity: sha512-EykJT/Q1KjTWctppgIAgfSO0tKVuZUjhgMr17kqTumMl6Afv3EISleU7qZUzoXDFTAHTDC4NOoG/ZxU3EvlMPQ==} engines: {node: '>=8'} + has-symbols@1.1.0: + resolution: {integrity: sha512-1cDNdwJ2Jaohmb3sg4OmKaMBwuC48sYni5HUw2DvsC8LjGTLK9h+eb1X6RyuOHe4hT0ULCW68iomhjUoKUqlPQ==} + engines: {node: '>= 0.4'} + + hasown@2.0.4: + resolution: {integrity: sha512-T2UbfbBEF32wiepXIsMlTW9+dDYC6wMh/t/vYA4tuOMKqWz/n3vr1NFSxQiyP+zk2mXsoMA/i/7qV6LKut1t1A==} + engines: {node: '>= 0.4'} + + hono@4.12.28: + resolution: {integrity: sha512-YwUvVpSF7m1yOblFPrU3Hbo8XhPheBoiyfGuII6z19LnOr6JpDnyyp7LFNrfV56wS8tpvtBFGRISHN02pDdLOA==} + engines: {node: '>=16.9.0'} + hookable@6.1.1: resolution: {integrity: sha512-U9LYDy1CwhMCnprUfeAZWZGByVbhd54hwepegYTK7Pi5NvqEj63ifz5z+xukznehT7i6NIZRu89Ay1AZmRsLEQ==} @@ -3014,6 +3198,10 @@ packages: html-escaper@2.0.2: resolution: {integrity: sha512-H2iMtd0I4Mt5eYiapRdIDjp+XzelXQ0tFE4JS7YFwFevXXMmOp9myNrUvCg0D6ws8iqkRPBfKHgbwig1SmlLfg==} + http-errors@2.0.1: + resolution: {integrity: sha512-4FbRdAX+bSdmo4AUFuS0WNiPz8NgFt+r8ThgNWmlrjQjt1Q7ZR9+zTlce2859x4KSXrwIsaeTqDoKQmtP8pLmQ==} + engines: {node: '>= 0.8'} + http-proxy-agent@7.0.2: resolution: {integrity: sha512-T1gkAiYYDWYx3V5Bmyu7HcfcvL7mUrTWiM6yOfa3PIphViJ/gFPbvidQ+veqSOHci/PxBcDabeUNCzpOODJZig==} engines: {node: '>= 14'} @@ -3026,6 +3214,10 @@ packages: resolution: {integrity: sha512-4fCk79wshMdzMp2rH06qWrJE4iolqLhCUH+OiuIgU++RB0+94NlDL81atO7GX55uUKueo0txHNtvEyI6D7WdMw==} engines: {node: '>=0.10.0'} + iconv-lite@0.7.3: + resolution: {integrity: sha512-IKXpvIzjnC9XTAUbVBcMfGS0EPaIXtW6v+zr+RRp+hqULEpo0owZax6wyRwPOJbWbzjYspQwusTsfVr0ifh4uQ==} + engines: {node: '>=0.10.0'} + ignore@5.3.2: resolution: {integrity: sha512-hsBTNUqQTDwkWtcdYI2i06Y/nUBEsNEDJKjWdigLvegy8kDuJAS8uRlpkkcQpyEXL0Z/pjDy5HBmMjRCJ2gq+g==} engines: {node: '>= 4'} @@ -3045,6 +3237,9 @@ packages: resolution: {integrity: sha512-JmXMZ6wuvDmLiHEml9ykzqO6lwFbof0GG4IkcGaENdCRDDmMVnny7s5HsIgHCbaq0w2MyPhDqkhTUgS2LU2PHA==} engines: {node: '>=0.8.19'} + inherits@2.0.4: + resolution: {integrity: sha512-k/vGaX4/Yla3WzyMCvTQOXYeIHvqOKtnqBduzTHpzpQZzAskKMhZ2K+EnBiSM9zGSoIFeMpXKxa4dYeZIQqewQ==} + internmap@1.0.1: resolution: {integrity: sha512-lDB5YccMydFBtasVtxnZ3MRBHuaoE8GKsppq+EchKL2U4nK/DmEpPHNH8MZe5HkMtpSiTSOZwfN0tzYjO/lJEw==} @@ -3052,6 +3247,14 @@ packages: resolution: {integrity: sha512-5Hh7Y1wQbvY5ooGgPbDaL5iYLAPzMTUrjMulskHLH6wnv/A+1q5rgEaiuqEjB+oxGXIVZs1FF+R/KPN3ZSQYYg==} engines: {node: '>=12'} + ip-address@10.2.0: + resolution: {integrity: sha512-/+S6j4E9AHvW9SWMSEY9Xfy66O5PWvVEJ08O0y5JGyEKQpojb0K0GKpz/v5HJ/G0vi3D2sjGK78119oXZeE0qA==} + engines: {node: '>= 12'} + + ipaddr.js@1.9.1: + resolution: {integrity: sha512-0KI/607xoxSToH7GjN1FfSbLoU0+btTicjsQSWQlh/hZykN8KpmMf7uYwPW3R+akZ6R/w18ZlXSHBYXiYUPO3g==} + engines: {node: '>= 0.10'} + is-extglob@2.1.1: resolution: {integrity: sha512-SbKbANkN603Vi4jEZv49LeVJMn4yGwsbzZworEoyEiutsN3nJYdbO36zfhGJ6QEDpOZIFkDtnq5JRxmvl3jsoQ==} engines: {node: '>=0.10.0'} @@ -3063,6 +3266,9 @@ packages: is-potential-custom-element-name@1.0.1: resolution: {integrity: sha512-bCYeRA2rVibKZd+s2625gGnGF/t7DSqDs4dP7CrLA1m7jKWz6pps0LpYLJN8Q64HtmPKJ1hrN3nzPNKFEKOUiQ==} + is-promise@4.0.0: + resolution: {integrity: sha512-hvpoI6korhJMnej285dSg6nu1+e6uxs7zG3BYAm5byqDsgJNWwxzM6z6iZiAgQR4TJ30JmBTOwqZUw3WlyH3AQ==} + isexe@2.0.0: resolution: {integrity: sha512-RHxMLp9lnKHGHRng9QFhRCMbYAcVpn69smSGcq3f36xjgVVWThj4qqLbTLlq7Ssj8B+fIQ1EuCEGI2lKsyQeIw==} @@ -3082,6 +3288,9 @@ packages: resolution: {integrity: sha512-AC/7JofJvZGrrneWNaEnJeOLUx+JlGt7tNa0wZiRPT4MY1wmfKjt2+6O2p2uz2+skll8OZZmJMNqeke7kKbNgQ==} hasBin: true + jose@6.2.3: + resolution: {integrity: sha512-YYVDInQKFJfR/xa3ojUTl8c2KoTwiL1R5Wg9YCydwH0x0B9grbzlg5HC7mMjCtUJjbQ/YnGEZIhI5tCgfTb4Hw==} + js-tokens@10.0.0: resolution: {integrity: sha512-lM/UBzQmfJRo9ABXbPWemivdCW8V2G8FHaHdypQaIy523snUjog0W71ayWXTjiR+ixeMyVHN2XcpnTd/liPg/Q==} @@ -3119,6 +3328,12 @@ packages: json-schema-traverse@0.4.1: resolution: {integrity: sha512-xbbCH5dCYU5T8LcEhhuh7HJ88HXuW3qsI3Y0zOZFKfZEHcpWiHU/Jxzk629Brsab/mMiHQti9wMP+845RPe3Vg==} + json-schema-traverse@1.0.0: + resolution: {integrity: sha512-NM8/P9n3XjXhIZn1lLhkFaACTOURQXjWhV4BA/RnOv8xvgqtqpAX9IO4mRQxSx1Rlo4tqzeqb0sOlruaOy3dug==} + + json-schema-typed@8.0.2: + resolution: {integrity: sha512-fQhoXdcvc3V28x7C7BMs4P5+kNlgUURe2jmUT1T//oBRMDrqy1QPelJimwZGo7Hg9VPV3EQV5Bnq4hbFy2vetA==} + json-stable-stringify-without-jsonify@1.0.1: resolution: {integrity: sha512-Bdboy+l7tA3OGW6FjyFHWkP5LuByj1Tk33Ljyq0axyzdk9//JSi2u3fP1QSmd1KNwq6VOKYGlAu87CisVir6Pw==} @@ -3316,6 +3531,10 @@ packages: engines: {node: '>= 20'} hasBin: true + math-intrinsics@1.1.0: + resolution: {integrity: sha512-/IXtbwEk5HTPyEwyKX6hGkYXxM9nbj64B+ilVJnC/R6B0pH5G4V3b0pVbL7DBj4tkhBAppbQUlf6F6Xl9LHu1g==} + engines: {node: '>= 0.4'} + mdast-util-find-and-replace@3.0.2: resolution: {integrity: sha512-Tmd1Vg/m3Xz43afeNxDIhWRtFZgM2VLyaf4vSTYwudTyeuTneoL3qtWMA5jeLyz/O1vDJmmV4QuScFCA2tBPwg==} @@ -3352,6 +3571,14 @@ packages: mdn-data@2.27.1: resolution: {integrity: sha512-9Yubnt3e8A0OKwxYSXyhLymGW4sCufcLG6VdiDdUGVkPhpqLxlvP5vl1983gQjJl3tqbrM731mjaZaP68AgosQ==} + media-typer@1.1.0: + resolution: {integrity: sha512-aisnrDP4GNe06UcKFnV5bfMNPBUw4jsLGaWwWfnH3v02GnBuXX2MCVn5RbrWo0j3pczUilYblq7fQ7Nw2t5XKw==} + engines: {node: '>= 0.8'} + + merge-descriptors@2.0.0: + resolution: {integrity: sha512-Snk314V5ayFLhp3fkUREub6WtjBfPdCPY1Ln8/8munuLuiYhsABgBVWsozAG+MWMbVEvcdcpbi9R7ww22l9Q3g==} + engines: {node: '>=18'} + mermaid@11.16.0: resolution: {integrity: sha512-Zvm3kbstgdpvIJPPItlL7fppIZ3kibvc1oZIGxdvk9t6UFz6flv+Jw7FtRGKwfcI8OckmH04LqG6LlS6X4B1pA==} @@ -3439,6 +3666,14 @@ packages: micromark@4.0.2: resolution: {integrity: sha512-zpe98Q6kvavpCr1NPVSCMebCKfD7CA2NqZ+rykeNhONIJBpc1tFKt9hucLGwha3jNTNI8lHpctWJWoimVF4PfA==} + mime-db@1.54.0: + resolution: {integrity: sha512-aU5EJuIN2WDemCcAp2vFBfp/m4EAhWJnUNSSw0ixs7/kXbd6Pg64EmwJkNdFhB8aWt1sH2CTXrLxo/iAGV3oPQ==} + engines: {node: '>= 0.6'} + + mime-types@3.0.2: + resolution: {integrity: sha512-Lbgzdk0h4juoQ9fCKXW4by0UJqj+nOOrI9MJ1sSj4nI8aI2eo1qmvQEie4VD1glsS250n15LsWsYtCugiStS5A==} + engines: {node: '>=18'} + minimatch@10.2.5: resolution: {integrity: sha512-MULkVLfKGYDFYejP07QOurDLLQpcjk7Fw+7jXS2R2czRQzR56yHRveU5NDJEOviH+hETZKSkIk5c+T23GjFUMg==} engines: {node: 18 || 20 || >=22} @@ -3458,6 +3693,10 @@ packages: natural-compare@1.4.0: resolution: {integrity: sha512-OWND8ei3VtNC9h7V60qff3SVobHr996CTwgxubgyQYEpg290h9J0buyECNNJexkFm5sOajh5G116RYA1c8ZMSw==} + negotiator@1.0.0: + resolution: {integrity: sha512-8Ofs/AUQh8MaEcrlq5xOX0CQ9ypTF5dl78mjlMNfOK08fzpgTHQRQPBxcPlEtIw0yRpws+Zo/3r+5WRby7u3Gg==} + engines: {node: '>= 0.6'} + node-domexception@1.0.0: resolution: {integrity: sha512-/jKZoMpw0F8GRwl4/eLROPA3cfcXtLApP0QzLmUT/HuPCZWyB7IY9ZrMeKw2O/nFIqPQB3PVM9aYm0F312AXDQ==} engines: {node: '>=10.5.0'} @@ -3467,10 +3706,25 @@ packages: resolution: {integrity: sha512-dRB78srN/l6gqWulah9SrxeYnxeddIG30+GOqK/9OlLVyLg3HPnr6SqOWTWOXKRwC2eGYCkZ59NNuSgvSrpgOA==} engines: {node: ^12.20.0 || ^14.13.1 || >=16.0.0} + object-assign@4.1.1: + resolution: {integrity: sha512-rJgTQnkUnH1sFw8yT6VSU3zD3sWmu6sZhIseY8VX+GRu3P6F7Fu+JNDoXfklElbLJSnc3FUQHVe4cU5hj+BcUg==} + engines: {node: '>=0.10.0'} + + object-inspect@1.13.4: + resolution: {integrity: sha512-W67iLl4J2EXEGTbfeHCffrjDfitvLANg0UlX3wFUUSTx92KXRFegMHUVgSqE+wvhAbi4WqjGg9czysTV2Epbew==} + engines: {node: '>= 0.4'} + obug@2.1.3: resolution: {integrity: sha512-9miFgM2OFba7hB+pRgvtV84pYTBaoTHohvmIgiRt6dRIzbwEOIaNaP+dIlGs2fNFoB0SeISs0Jz5WFVRid6Xyg==} engines: {node: '>=12.20.0'} + on-finished@2.4.1: + resolution: {integrity: sha512-oVlzkg3ENAhCk2zdv7IJwd/QUD4z2RxRwpkcGY8psCVcCYZNq4wYnVWALHM+brtuJjePWiYF/ClmuDr8Ch5+kg==} + engines: {node: '>= 0.8'} + + once@1.4.0: + resolution: {integrity: sha512-lNaJgI+2Q5URQBkccEKHTQOPaXdUxnZZElQTZY0MFUAuaEqe1E+Nyvgdz/aIyNi6Z9MzO5dv1H8n58/GELp3+w==} + openai@6.26.0: resolution: {integrity: sha512-zd23dbWTjiJ6sSAX6s0HrCZi41JwTA1bQVs0wLQPZ2/5o2gxOJA5wh7yOAUgwYybfhDXyhwlpeQf7Mlgx8EOCA==} hasBin: true @@ -3512,6 +3766,10 @@ packages: parse5@8.0.1: resolution: {integrity: sha512-z1e/HMG90obSGeidlli3hj7cbocou0/wa5HacvI3ASx34PecNjNQeaHNo5WIZpWofN9kgkqV1q5YvXe3F0FoPw==} + parseurl@1.3.3: + resolution: {integrity: sha512-CiyeOxFT/JZyN5m0z9PfXw4SCBJ6Sygz1Dpl0wqjlhDEGGBP1GnsUVEL0p63hoG1fcj3fHynXi9NYO4nWOL+qQ==} + engines: {node: '>= 0.8'} + partial-json@0.1.7: resolution: {integrity: sha512-Njv/59hHaokb/hRUjce3Hdv12wd60MtM9Z5Olmn+nehe0QDAsRtRbJPvJ0Z91TusF0SuZRIvnM+S4l6EIP8leA==} @@ -3530,6 +3788,9 @@ packages: resolution: {integrity: sha512-ojmeN0qd+y0jszEtoY48r0Peq5dwMEkIlCOu6Q5f41lfkswXuKtYrhgoTpLnyIcHm24Uhqx+5Tqm2InSwLhE6Q==} engines: {node: '>=8'} + path-to-regexp@8.4.2: + resolution: {integrity: sha512-qRcuIdP69NPm4qbACK+aDogI5CBDMi1jKe0ry5rSQJz8JVLsC7jV8XpiJjGRLLol3N+R5ihGYcrPLTno6pAdBA==} + pathe@2.0.3: resolution: {integrity: sha512-WUjGcAqP1gQacoQe+OBJsFA7Ld4DyXuUIjZ5cc75cLHvJ7dtNsTugphxIADwspS+AraAUePCKrSVtPLFj/F88w==} @@ -3540,6 +3801,10 @@ packages: resolution: {integrity: sha512-QP88BAKvMam/3NxH6vj2o21R6MjxZUAd6nlwAS/pnGvN9IVLocLHxGYIzFhg6fUQ+5th6P4dv4eW9jX3DSIj7A==} engines: {node: '>=12'} + pkce-challenge@5.0.1: + resolution: {integrity: sha512-wQ0b/W4Fr01qtpHlqSqspcj3EhBvimsdh0KlHhH8HRZnMsEa0ea2fTULOXOS9ccQr3om+GcGRk4e+isrZWV8qQ==} + engines: {node: '>=16.20.0'} + points-on-curve@0.2.0: resolution: {integrity: sha512-0mYKnYYe9ZcqMCWhUjItv/oHjvgEsfKvnUTg8sAtnHr3GVy7rGkXCb6d5cSyqrWqL4k81b9CPg3urd+T7aop3A==} @@ -3558,6 +3823,10 @@ packages: resolution: {integrity: sha512-RJJPTTpvFfHcWLkIa2JFWK4XvtSzS0yEWDmunqHXli1h3JlkbcQZXDZdcWxv+JK3Xsl5/UFDPZ0iGm7DAengYw==} engines: {node: '>=12.0.0'} + proxy-addr@2.0.7: + resolution: {integrity: sha512-llQsMLSUDUPT44jdrU/O37qlnifitDP+ZwrmmZcoSKyLKvtZxpyV0n2/bD/N4tBAAZ/gJEdZU7KMraoK1+XYAg==} + engines: {node: '>= 0.10'} + publint@0.3.21: resolution: {integrity: sha512-OqejcnMV6E9zel2oCrUOJEiiFkGiAAni0A6ibfQNh1k9Gu5z4F+Yso8lllam7AzmV6Do0vp7u3UpZNRBwuXaHQ==} engines: {node: '>=18'} @@ -3570,9 +3839,21 @@ packages: pure-rand@8.4.0: resolution: {integrity: sha512-IoM8YF/jY0hiugFo/wOWqfmarlE6J0wc6fDK1PhftMk7MGhVZl88sZimmqBBFomLOCSmcCCpsfj7wXASCpvK9A==} + qs@6.15.3: + resolution: {integrity: sha512-O9gl3zCl5h5blw1KGUzQKhA5oUXSl8rwUIM5o0S3nCXMliSvy5Dzx7/DJcI+SwgICv+IneSZwhBh1oSyEHA71A==} + engines: {node: '>=0.6'} + quansync@1.0.0: resolution: {integrity: sha512-5xZacEEufv3HSTPQuchrvV6soaiACMFnq1H8wkVioctoH3TRha9Sz66lOxRwPK/qZj7HPiSveih9yAyh98gvqA==} + range-parser@1.3.0: + resolution: {integrity: sha512-hek2mFQpPuI4E1BBKrSto+BU3e3x4xuarsbiwr3+lf7p44juvFMV0XFWQAP3xUyqXA4RrXLIoaSUGbSt056ZMw==} + engines: {node: '>= 0.6'} + + raw-body@3.0.2: + resolution: {integrity: sha512-K5zQjDllxWkf7Z5xJdV0/B0WTNqx6vxG70zJE4N0kBs4LovmEYWJzQGxC9bS9RAKu3bgM40lrd5zoLJ12MQ5BA==} + engines: {node: '>= 0.10'} + readdirp@4.1.2: resolution: {integrity: sha512-GDhwkLfywWL2s6vEjyhri+eXmfH6j1L7JE27WhqLeYzoh/A3DBaYGEj2H/HFZCn/kMfim73FXxEJTw06WtxQwg==} engines: {node: '>= 14.18.0'} @@ -3623,6 +3904,10 @@ packages: roughjs@4.6.6: resolution: {integrity: sha512-ZUz/69+SYpFN/g/lUlo2FXcIjRkSu3nDarreVdGGndHEBJ6cXPdKguS8JGxwj5HA5xIbVKSmLgr5b3AWxtRfvQ==} + router@2.2.0: + resolution: {integrity: sha512-nLTrUKm2UyiL7rlhapu/Zl45FwNgkZGaCpZbIHajDYgwlJCOzLSk+cIPAnsEqV955GjILJnKbdQC1nVPz+gAYQ==} + engines: {node: '>= 18'} + rw@1.3.3: resolution: {integrity: sha512-PdhdWy89SiZogBLaw42zdeqtRJ//zFd2PgQavcICDUgJT5oW10QCRKbJ6bg4r0/UY2M6BWd5tkxuGFRvCkgfHQ==} @@ -3648,6 +3933,17 @@ packages: engines: {node: '>=10'} hasBin: true + send@1.2.1: + resolution: {integrity: sha512-1gnZf7DFcoIcajTjTwjwuDjzuz4PPcY2StKPlsGAQ1+YH20IRVrBaXSWmdjowTJ6u8Rc01PoYOGHXfP1mYcZNQ==} + engines: {node: '>= 18'} + + serve-static@2.2.1: + resolution: {integrity: sha512-xRXBn0pPqQTVQiC8wyQrKs2MOlX24zQ0POGaj0kultvoOCstBQM5yvOhAVSUwOMjQtTvsPWoNCHfPGwaaQJhTw==} + engines: {node: '>= 18'} + + setprototypeof@1.2.0: + resolution: {integrity: sha512-E5LDX7Wrp85Kil5bhZv46j8jOeboKq5JMmYM3gVGdGH8xFpPWXUMsNrlODCrkoxMEeNi/XZIwuRvY4XNwYMJpw==} + shebang-command@2.0.0: resolution: {integrity: sha512-kHxr2zZpYtdmrN1qDjrrX/Z1rR1kG8Dx+gkpK1G4eXmvXswmcE1hTWBWYUzlraYw1/yZp6YuDY77YtvbN0dmDA==} engines: {node: '>=8'} @@ -3656,6 +3952,22 @@ packages: resolution: {integrity: sha512-7++dFhtcx3353uBaq8DDR4NuxBetBzC7ZQOhmTQInHEd6bSrXdiEyzCvG07Z44UYdLShWUyXt5M/yhz8ekcb1A==} engines: {node: '>=8'} + side-channel-list@1.0.1: + resolution: {integrity: sha512-mjn/0bi/oUURjc5Xl7IaWi/OJJJumuoJFQJfDDyO46+hBWsfaVM65TBHq2eoZBhzl9EchxOijpkbRC8SVBQU0w==} + engines: {node: '>= 0.4'} + + side-channel-map@1.0.1: + resolution: {integrity: sha512-VCjCNfgMsby3tTdo02nbjtM/ewra6jPHmpThenkTYh8pG9ucZ/1P8So4u4FGBek/BjpOVsDCMoLA/iuBKIFXRA==} + engines: {node: '>= 0.4'} + + side-channel-weakmap@1.0.2: + resolution: {integrity: sha512-WPS/HvHQTYnHisLo9McqBHOJk2FkHO/tlpvldyrnem4aeQp4hai3gythswg6p01oSoTl58rcpiFAjF2br2Ak2A==} + engines: {node: '>= 0.4'} + + side-channel@1.1.1: + resolution: {integrity: sha512-6x6dK6zJdpTzF4sQeNYxwtvBzf6Eg4GtlesS94HOvTudUeyK2WXAaIfmDgsyslYrRBeFIlsi54AYsFGUuhmvrQ==} + engines: {node: '>= 0.4'} + siginfo@2.0.0: resolution: {integrity: sha512-ybx0WO1/8bSBLEWXZvEd7gMW3Sn3JFlW3TvX1nREbDLRNQNaeNN8WK0meBwPdAaOI7TtRRRJn/Es1zhrrCHu7g==} @@ -3670,6 +3982,10 @@ packages: stackback@0.0.2: resolution: {integrity: sha512-1XMJE5fQo1jGH6Y/7ebnwPOBEkIEnT4QF32d5R1+VXdXveM0IBMJt8zfaxX1P3QhVwrYe+576+jkANtSS2mBbw==} + statuses@2.0.2: + resolution: {integrity: sha512-DvEy55V3DB7uknRo+4iOGT5fP1slR8wQohVdknigZPMpMstaKJQWhwiYBACJE3Ul2pTnATihhBYnRhZQHGBiRw==} + engines: {node: '>= 0.8'} + std-env@4.1.0: resolution: {integrity: sha512-Rq7ybcX2RuC55r9oaPVEW7/xu3tj8u4GeBYHBWCychFtzMIr86A7e3PPEBPT37sHStKX3+TiX/Fr/ACmJLVlLQ==} @@ -3716,6 +4032,10 @@ packages: resolution: {integrity: sha512-RfEzKWcq5fHUOFq7J3rl3Oz6ylKGtcHqUznzj4EcXsxLSIjJcvpbXAQtWGeJQ0xKnimR5e0Cn+cn9TssfMzm+g==} hasBin: true + toidentifier@1.0.1: + resolution: {integrity: sha512-o5sSPKEkg/DIQNmH43V0/uerLrpzVedkUh8tGNvaeXpfpuwjKenlSox/2O/BTlZUtEe+JG7s5YhEz608PlAHRA==} + engines: {node: '>=0.6'} + tough-cookie@6.0.1: resolution: {integrity: sha512-LktZQb3IeoUWB9lqR5EWTHgW/VTITCXg4D21M+lvybRVdylLrRMnqaIONLVb5mav8vM19m44HIcGq4qASeu2Qw==} engines: {node: '>=16'} @@ -3798,6 +4118,10 @@ packages: resolution: {integrity: sha512-XleUoc9uwGXqjWwXaUTZAmzMcFZ5858QA2vvx1Ur5xIcixXIP+8LnFDgRplU30us6teqdlskFfu+ae4K79Ooew==} engines: {node: '>= 0.8.0'} + type-is@2.1.0: + resolution: {integrity: sha512-faYHw0anBbc/kWF3zFTEnxSFOAGUX9GFbOBthvDdLsIlEoWOFOtS0zgCiQYwIskL9iGXZL3kAXD8OoZ4GmMATA==} + engines: {node: '>= 18'} + typebox@1.1.38: resolution: {integrity: sha512-pZ0aQPmMmXoUvSbeuWf/Hzsc+avNw/Zd6VeE8CFgkVGWyuHPJvqeJJDeJqLve+K70LvjYIoleGcoJHPT17cWoA==} @@ -3839,6 +4163,10 @@ packages: unist-util-visit@5.1.0: resolution: {integrity: sha512-m+vIdyeCOpdr/QeQCu2EzxX/ohgS8KbnPDgFni4dQsfSCtpz8UqDyY5GjRru8PDKuYn7Fq19j1CQ+nJSsGKOzg==} + unpipe@1.0.0: + resolution: {integrity: sha512-pjy2bYhSsufwWlKwPc+l3cN7+wuJlK6uz0YdJEOlQDbl6jo/YlPi4mb8agUkVC8BF7V8NuzeyPNqRksA3hztKQ==} + engines: {node: '>= 0.8'} + uri-js@4.4.1: resolution: {integrity: sha512-7rKUyy33Q1yc98pQ1DAmLtwX109F7TIfWlW1Ydo8Wl1ii1SeHieeh0HHfPeL2fMXK6z0s8ecKs9frCuLJvndBg==} @@ -3846,6 +4174,10 @@ packages: resolution: {integrity: sha512-6ZxzVpzDXDa3bJWaHilVayA+BH/1zmxCJoVgvmqJnid/gPoKHxUrS/aC/T6LGQtNHT+XHG9fXPJB4d+IrU30Ew==} hasBin: true + vary@1.1.2: + resolution: {integrity: sha512-BNGbWLfd0eUPabhkXUVm0j8uuvREyTh5ovRa/dyow/BqAbZJyC+5fU+IzQOzmAKzYqYRAISoRhdQr3eIZ/PXqg==} + engines: {node: '>= 0.8'} + vite-tsconfig-paths@6.1.1: resolution: {integrity: sha512-2cihq7zliibCCZ8P9cKJrQBkfgdvcFkOOc3Y02o3GWUDLgqjWsZudaoiuOwO/gzTzy17cS5F7ZPo4bsnS4DGkg==} peerDependencies: @@ -3973,6 +4305,9 @@ packages: resolution: {integrity: sha512-BN22B5eaMMI9UMtjrGd5g5eCYPpCPDUy0FJXbYsaT5zYxjFOckS53SQDE3pWkVoWpHXVb3BrYcEN4Twa55B5cA==} engines: {node: '>=0.10.0'} + wrappy@1.0.2: + resolution: {integrity: sha512-l4Sp/DRseor9wL6EvV2+TuQn63dMkPjZ/sp9XkghTEbV9KlPS1xUsZ3u7/IQO4wxtcFB4bgpQPRcR3QCvezPcQ==} + ws@8.21.0: resolution: {integrity: sha512-Vsp28b7DRcimFQvrqu2Wek3z1iYxDCWqHYB8Qsnk/S4RfaCQzPGPyBNuVjJV3cd6UiKtUtp6sNM77gWvzcCH+g==} engines: {node: '>=10.0.0'} @@ -4370,11 +4705,11 @@ snapshots: '@csstools/css-tokenizer@4.0.0': {} - '@earendil-works/pi-ai@0.79.3(ws@8.21.0)(zod@4.4.3)': + '@earendil-works/pi-ai@0.79.3(@modelcontextprotocol/sdk@1.29.0(zod@4.4.3))(ws@8.21.0)(zod@4.4.3)': dependencies: '@anthropic-ai/sdk': 0.91.1(zod@4.4.3) '@aws-sdk/client-bedrock-runtime': 3.1048.0 - '@google/genai': 1.52.0 + '@google/genai': 1.52.0(@modelcontextprotocol/sdk@1.29.0(zod@4.4.3)) '@mistralai/mistralai': 2.2.1 '@smithy/node-http-handler': 4.7.3 http-proxy-agent: 7.0.2 @@ -4532,17 +4867,23 @@ snapshots: '@exodus/bytes@1.15.1': {} - '@google/genai@1.52.0': + '@google/genai@1.52.0(@modelcontextprotocol/sdk@1.29.0(zod@4.4.3))': dependencies: google-auth-library: 10.7.0 p-retry: 4.6.2 protobufjs: 7.6.4 ws: 8.21.0 + optionalDependencies: + '@modelcontextprotocol/sdk': 1.29.0(zod@4.4.3) transitivePeerDependencies: - bufferutil - supports-color - utf-8-validate + '@hono/node-server@1.19.14(hono@4.12.28)': + dependencies: + hono: 4.12.28 + '@humanfs/core@0.19.2': dependencies: '@humanfs/types': 0.15.0 @@ -4594,6 +4935,28 @@ snapshots: - bufferutil - utf-8-validate + '@modelcontextprotocol/sdk@1.29.0(zod@4.4.3)': + dependencies: + '@hono/node-server': 1.19.14(hono@4.12.28) + ajv: 8.20.0 + ajv-formats: 3.0.1(ajv@8.20.0) + content-type: 1.0.5 + cors: 2.8.6 + cross-spawn: 7.0.6 + eventsource: 3.0.7 + eventsource-parser: 3.1.0 + express: 5.2.1 + express-rate-limit: 8.5.2(express@5.2.1) + hono: 4.12.28 + jose: 6.2.3 + json-schema-typed: 8.0.2 + pkce-challenge: 5.0.1 + raw-body: 3.0.2 + zod: 4.4.3 + zod-to-json-schema: 3.25.2(zod@4.4.3) + transitivePeerDependencies: + - supports-color + '@napi-rs/wasm-runtime@1.1.5(@emnapi/core@1.10.0)(@emnapi/runtime@1.10.0)': dependencies: '@emnapi/core': 1.10.0 @@ -5255,6 +5618,11 @@ snapshots: convert-source-map: 2.0.0 tinyrainbow: 3.1.0 + accepts@2.0.0: + dependencies: + mime-types: 3.0.2 + negotiator: 1.0.0 + acorn-jsx@5.3.2(acorn@8.17.0): dependencies: acorn: 8.17.0 @@ -5263,6 +5631,10 @@ snapshots: agent-base@7.1.4: {} + ajv-formats@3.0.1(ajv@8.20.0): + optionalDependencies: + ajv: 8.20.0 + ajv@6.15.0: dependencies: fast-deep-equal: 3.1.3 @@ -5270,6 +5642,13 @@ snapshots: json-schema-traverse: 0.4.1 uri-js: 4.4.1 + ajv@8.20.0: + dependencies: + fast-deep-equal: 3.1.3 + fast-uri: 3.1.3 + json-schema-traverse: 1.0.0 + require-from-string: 2.0.2 + ansis@4.3.1: {} anynum@1.0.0: {} @@ -5302,6 +5681,20 @@ snapshots: birpc@4.0.0: {} + body-parser@2.3.0: + dependencies: + bytes: 3.1.2 + content-type: 2.0.0 + debug: 4.4.3 + http-errors: 2.0.1 + iconv-lite: 0.7.3 + on-finished: 2.4.1 + qs: 6.15.3 + raw-body: 3.0.2 + type-is: 2.1.0 + transitivePeerDependencies: + - supports-color + bowser@2.14.1: {} brace-expansion@5.0.6: @@ -5310,8 +5703,20 @@ snapshots: buffer-equal-constant-time@1.0.1: {} + bytes@3.1.2: {} + cac@7.0.0: {} + call-bind-apply-helpers@1.0.2: + dependencies: + es-errors: 1.3.0 + function-bind: 1.1.2 + + call-bound@1.0.4: + dependencies: + call-bind-apply-helpers: 1.0.2 + get-intrinsic: 1.3.0 + ccount@2.0.1: {} chai@6.2.2: {} @@ -5326,8 +5731,18 @@ snapshots: commander@8.3.0: {} + content-disposition@1.1.0: {} + + content-type@1.0.5: {} + + content-type@2.0.0: {} + convert-source-map@2.0.0: {} + cookie-signature@1.2.2: {} + + cookie@0.7.2: {} + cordis@4.0.0-rc.6(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.4): dependencies: '@standard-schema/spec': 1.1.0 @@ -5344,6 +5759,11 @@ snapshots: '@cordisjs/plugin-include': link:vendor/include '@cordisjs/plugin-loader': link:vendor/loader + cors@2.8.6: + dependencies: + object-assign: 4.1.1 + vary: 1.1.2 + cose-base@1.0.3: dependencies: layout-base: 1.0.2 @@ -5578,6 +5998,8 @@ snapshots: dependencies: robust-predicates: 3.0.3 + depd@2.0.0: {} + dequal@2.0.3: {} detect-libc@2.1.2: {} @@ -5596,16 +6018,34 @@ snapshots: optionalDependencies: oxc-resolver: 11.20.0 + dunder-proto@1.0.1: + dependencies: + call-bind-apply-helpers: 1.0.2 + es-errors: 1.3.0 + gopd: 1.2.0 + ecdsa-sig-formatter@1.0.11: dependencies: safe-buffer: 5.2.1 + ee-first@1.1.1: {} + empathic@2.0.1: {} + encodeurl@2.0.0: {} + entities@8.0.0: {} + es-define-property@1.0.1: {} + + es-errors@1.3.0: {} + es-module-lexer@2.1.0: {} + es-object-atoms@1.1.2: + dependencies: + es-errors: 1.3.0 + es-toolkit@1.49.0: {} esbuild@0.28.1: @@ -5637,6 +6077,8 @@ snapshots: '@esbuild/win32-ia32': 0.28.1 '@esbuild/win32-x64': 0.28.1 + escape-html@1.0.3: {} + escape-string-regexp@4.0.0: {} escape-string-regexp@5.0.0: {} @@ -5719,8 +6161,54 @@ snapshots: esutils@2.0.3: {} + etag@1.8.1: {} + + eventsource-parser@3.1.0: {} + + eventsource@3.0.7: + dependencies: + eventsource-parser: 3.1.0 + expect-type@1.3.0: {} + express-rate-limit@8.5.2(express@5.2.1): + dependencies: + express: 5.2.1 + ip-address: 10.2.0 + + express@5.2.1: + dependencies: + accepts: 2.0.0 + body-parser: 2.3.0 + content-disposition: 1.1.0 + content-type: 1.0.5 + cookie: 0.7.2 + cookie-signature: 1.2.2 + debug: 4.4.3 + depd: 2.0.0 + encodeurl: 2.0.0 + escape-html: 1.0.3 + etag: 1.8.1 + finalhandler: 2.1.1 + fresh: 2.0.0 + http-errors: 2.0.1 + merge-descriptors: 2.0.0 + mime-types: 3.0.2 + on-finished: 2.4.1 + once: 1.4.0 + parseurl: 1.3.3 + proxy-addr: 2.0.7 + qs: 6.15.3 + range-parser: 1.3.0 + router: 2.2.0 + send: 1.2.1 + serve-static: 2.2.1 + statuses: 2.0.2 + type-is: 2.1.0 + vary: 1.1.2 + transitivePeerDependencies: + - supports-color + extend@3.0.2: {} fast-check@4.8.0: @@ -5733,6 +6221,8 @@ snapshots: fast-levenshtein@2.0.6: {} + fast-uri@3.1.3: {} + fast-xml-builder@1.2.0: dependencies: path-expression-matcher: 1.5.0 @@ -5762,6 +6252,17 @@ snapshots: dependencies: flat-cache: 4.0.1 + finalhandler@2.1.1: + dependencies: + debug: 4.4.3 + encodeurl: 2.0.0 + escape-html: 1.0.3 + on-finished: 2.4.1 + parseurl: 1.3.3 + statuses: 2.0.2 + transitivePeerDependencies: + - supports-color + find-up@5.0.0: dependencies: locate-path: 6.0.0 @@ -5782,9 +6283,15 @@ snapshots: dependencies: fetch-blob: 3.2.0 + forwarded@0.2.0: {} + + fresh@2.0.0: {} + fsevents@2.3.3: optional: true + function-bind@1.1.2: {} + gaxios@7.1.5: dependencies: extend: 3.0.2 @@ -5801,6 +6308,24 @@ snapshots: transitivePeerDependencies: - supports-color + get-intrinsic@1.3.0: + dependencies: + call-bind-apply-helpers: 1.0.2 + es-define-property: 1.0.1 + es-errors: 1.3.0 + es-object-atoms: 1.1.2 + function-bind: 1.1.2 + get-proto: 1.0.1 + gopd: 1.2.0 + has-symbols: 1.1.0 + hasown: 2.0.4 + math-intrinsics: 1.1.0 + + get-proto@1.0.1: + dependencies: + dunder-proto: 1.0.1 + es-object-atoms: 1.1.2 + get-tsconfig@4.14.0: dependencies: resolve-pkg-maps: 1.0.0 @@ -5828,10 +6353,20 @@ snapshots: google-logging-utils@1.1.3: {} + gopd@1.2.0: {} + hachure-fill@0.5.2: {} has-flag@4.0.0: {} + has-symbols@1.1.0: {} + + hasown@2.0.4: + dependencies: + function-bind: 1.1.2 + + hono@4.12.28: {} + hookable@6.1.1: {} html-encoding-sniffer@6.0.0: @@ -5842,6 +6377,14 @@ snapshots: html-escaper@2.0.2: {} + http-errors@2.0.1: + dependencies: + depd: 2.0.0 + inherits: 2.0.4 + setprototypeof: 1.2.0 + statuses: 2.0.2 + toidentifier: 1.0.1 + http-proxy-agent@7.0.2: dependencies: agent-base: 7.1.4 @@ -5860,6 +6403,10 @@ snapshots: dependencies: safer-buffer: 2.1.2 + iconv-lite@0.7.3: + dependencies: + safer-buffer: 2.1.2 + ignore@5.3.2: {} ignore@7.0.5: {} @@ -5870,10 +6417,16 @@ snapshots: imurmurhash@0.1.4: {} + inherits@2.0.4: {} + internmap@1.0.1: {} internmap@2.0.3: {} + ip-address@10.2.0: {} + + ipaddr.js@1.9.1: {} + is-extglob@2.1.1: {} is-glob@4.0.3: @@ -5882,6 +6435,8 @@ snapshots: is-potential-custom-element-name@1.0.1: {} + is-promise@4.0.0: {} + isexe@2.0.0: {} istanbul-lib-coverage@3.2.2: {} @@ -5899,6 +6454,8 @@ snapshots: jiti@2.7.0: {} + jose@6.2.3: {} + js-tokens@10.0.0: {} js-tokens@4.0.0: {} @@ -5948,6 +6505,10 @@ snapshots: json-schema-traverse@0.4.1: {} + json-schema-traverse@1.0.0: {} + + json-schema-typed@8.0.2: {} + json-stable-stringify-without-jsonify@1.0.1: {} jwa@2.0.1: @@ -6118,6 +6679,8 @@ snapshots: marked@16.4.2: {} + math-intrinsics@1.1.0: {} + mdast-util-find-and-replace@3.0.2: dependencies: '@types/mdast': 4.0.4 @@ -6222,6 +6785,10 @@ snapshots: mdn-data@2.27.1: {} + media-typer@1.1.0: {} + + merge-descriptors@2.0.0: {} + mermaid@11.16.0: dependencies: '@braintree/sanitize-url': 7.1.2 @@ -6437,6 +7004,12 @@ snapshots: transitivePeerDependencies: - supports-color + mime-db@1.54.0: {} + + mime-types@3.0.2: + dependencies: + mime-db: 1.54.0 + minimatch@10.2.5: dependencies: brace-expansion: 5.0.6 @@ -6449,6 +7022,8 @@ snapshots: natural-compare@1.4.0: {} + negotiator@1.0.0: {} + node-domexception@1.0.0: {} node-fetch@3.3.2: @@ -6457,8 +7032,20 @@ snapshots: fetch-blob: 3.2.0 formdata-polyfill: 4.0.10 + object-assign@4.1.1: {} + + object-inspect@1.13.4: {} + obug@2.1.3: {} + on-finished@2.4.1: + dependencies: + ee-first: 1.1.1 + + once@1.4.0: + dependencies: + wrappy: 1.0.2 + openai@6.26.0(ws@8.21.0)(zod@4.4.3): optionalDependencies: ws: 8.21.0 @@ -6539,6 +7126,8 @@ snapshots: dependencies: entities: 8.0.0 + parseurl@1.3.3: {} + partial-json@0.1.7: {} path-data-parser@0.1.0: {} @@ -6549,12 +7138,16 @@ snapshots: path-key@3.1.1: {} + path-to-regexp@8.4.2: {} + pathe@2.0.3: {} picocolors@1.1.1: {} picomatch@4.0.4: {} + pkce-challenge@5.0.1: {} + points-on-curve@0.2.0: {} points-on-path@0.2.1: @@ -6584,6 +7177,11 @@ snapshots: '@types/node': 25.9.3 long: 5.3.2 + proxy-addr@2.0.7: + dependencies: + forwarded: 0.2.0 + ipaddr.js: 1.9.1 + publint@0.3.21: dependencies: '@publint/pack': 0.1.4 @@ -6595,8 +7193,22 @@ snapshots: pure-rand@8.4.0: {} + qs@6.15.3: + dependencies: + es-define-property: 1.0.1 + side-channel: 1.1.1 + quansync@1.0.0: {} + range-parser@1.3.0: {} + + raw-body@3.0.2: + dependencies: + bytes: 3.1.2 + http-errors: 2.0.1 + iconv-lite: 0.7.3 + unpipe: 1.0.0 + readdirp@4.1.2: {} require-from-string@2.0.2: {} @@ -6672,6 +7284,16 @@ snapshots: points-on-curve: 0.2.0 points-on-path: 0.2.1 + router@2.2.0: + dependencies: + debug: 4.4.3 + depd: 2.0.0 + is-promise: 4.0.0 + parseurl: 1.3.3 + path-to-regexp: 8.4.2 + transitivePeerDependencies: + - supports-color + rw@1.3.3: {} sade@1.8.1: @@ -6693,12 +7315,67 @@ snapshots: semver@7.8.4: {} + send@1.2.1: + dependencies: + debug: 4.4.3 + encodeurl: 2.0.0 + escape-html: 1.0.3 + etag: 1.8.1 + fresh: 2.0.0 + http-errors: 2.0.1 + mime-types: 3.0.2 + ms: 2.1.3 + on-finished: 2.4.1 + range-parser: 1.3.0 + statuses: 2.0.2 + transitivePeerDependencies: + - supports-color + + serve-static@2.2.1: + dependencies: + encodeurl: 2.0.0 + escape-html: 1.0.3 + parseurl: 1.3.3 + send: 1.2.1 + transitivePeerDependencies: + - supports-color + + setprototypeof@1.2.0: {} + shebang-command@2.0.0: dependencies: shebang-regex: 3.0.0 shebang-regex@3.0.0: {} + side-channel-list@1.0.1: + dependencies: + es-errors: 1.3.0 + object-inspect: 1.13.4 + + side-channel-map@1.0.1: + dependencies: + call-bound: 1.0.4 + es-errors: 1.3.0 + get-intrinsic: 1.3.0 + object-inspect: 1.13.4 + + side-channel-weakmap@1.0.2: + dependencies: + call-bound: 1.0.4 + es-errors: 1.3.0 + get-intrinsic: 1.3.0 + object-inspect: 1.13.4 + side-channel-map: 1.0.1 + + side-channel@1.1.1: + dependencies: + es-errors: 1.3.0 + object-inspect: 1.13.4 + side-channel-list: 1.0.1 + side-channel-map: 1.0.1 + side-channel-weakmap: 1.0.2 + siginfo@2.0.0: {} smol-toml@1.6.1: {} @@ -6707,6 +7384,8 @@ snapshots: stackback@0.0.2: {} + statuses@2.0.2: {} + std-env@4.1.0: {} strip-json-comments@5.0.3: {} @@ -6742,6 +7421,8 @@ snapshots: dependencies: tldts-core: 7.4.5 + toidentifier@1.0.1: {} + tough-cookie@6.0.1: dependencies: tldts: 7.4.5 @@ -6803,6 +7484,12 @@ snapshots: dependencies: prelude-ls: 1.2.1 + type-is@2.1.0: + dependencies: + content-type: 2.0.0 + media-typer: 1.1.0 + mime-types: 3.0.2 + typebox@1.1.38: {} typescript-eslint@8.61.0(eslint@10.5.0(jiti@2.7.0))(typescript@6.0.3): @@ -6848,12 +7535,16 @@ snapshots: unist-util-is: 6.0.1 unist-util-visit-parents: 6.0.2 + unpipe@1.0.0: {} + uri-js@4.4.1: dependencies: punycode: 2.3.1 uuid@14.0.1: {} + vary@1.1.2: {} + vite-tsconfig-paths@6.1.1(typescript@6.0.3)(vite@8.0.16(@types/node@25.9.3)(esbuild@0.28.1)(jiti@2.7.0)(tsx@4.22.4)(yaml@2.9.0)): dependencies: debug: 4.4.3 @@ -6939,6 +7630,8 @@ snapshots: word-wrap@1.2.5: {} + wrappy@1.0.2: {} + ws@8.21.0: {} xml-name-validator@5.0.0: {} diff --git a/scripts/gen-config-catalog.ts b/scripts/gen-config-catalog.ts index eca36286ae..d260e7110b 100644 --- a/scripts/gen-config-catalog.ts +++ b/scripts/gen-config-catalog.ts @@ -537,6 +537,15 @@ function walkSchemaExpr( } return } + // A union of objects (discriminated union config): collect keys from all + // variants. Each variant is visited the same way as an intersect element. + if (method === 'union' && call.arguments[0] && ts.isArrayLiteralExpression(call.arguments[0])) { + for (const el of call.arguments[0].elements) { + const part = unwrapExpr(el) + if (ts.isCallExpression(part)) { visit(part); continue } + } + return + } // A chained refinement (`z.object({…}).default(…)` etc.): the keys live on // the call the chain hangs off — keep unwrapping toward it. const base = unwrapExpr(call.expression.expression) diff --git a/tsconfig.base.json b/tsconfig.base.json index 40e4dbe728..7cbb3acab2 100644 --- a/tsconfig.base.json +++ b/tsconfig.base.json @@ -52,6 +52,7 @@ "./packages/session-persistence/*/src", "./packages/ui/*/src", "./packages/util/*/src", + "./packages/mcp/*/src", "./packages/support/*/src" ] } diff --git a/tsconfig.build.json b/tsconfig.build.json index b6ba7901f2..1213898eec 100644 --- a/tsconfig.build.json +++ b/tsconfig.build.json @@ -54,6 +54,7 @@ { "path": "./packages/todo/tool-todo" }, { "path": "./packages/hooks/hook-protocol" }, { "path": "./packages/hooks/hooks-claude" }, - { "path": "./packages/hooks/hooks-codex" } + { "path": "./packages/hooks/hooks-codex" }, + { "path": "./packages/mcp/mcp-client" } ] } diff --git a/tsconfig.json b/tsconfig.json index 9cd7aa8a6d..77070ea044 100644 --- a/tsconfig.json +++ b/tsconfig.json @@ -65,6 +65,7 @@ { "path": "./packages/todo/tool-todo" }, { "path": "./packages/hooks/hook-protocol" }, { "path": "./packages/hooks/hooks-claude" }, - { "path": "./packages/hooks/hooks-codex" } + { "path": "./packages/hooks/hooks-codex" }, + { "path": "./packages/mcp/mcp-client" } ] } From f38111e5ca02c8723a68f4a7199f49a2945a1ac2 Mon Sep 17 00:00:00 2001 From: lintianle Date: Wed, 8 Jul 2026 11:58:12 +0800 Subject: [PATCH 02/26] fix: handle MCP transport disconnect and concurrent tool re-sync - Add client.onclose handler to unregister tools when the MCP server disconnects (crash or intentional close) - Replace bare `let disposers` with a shared mutable state object so the effect disposer and notification handler always reference the current generation - Serialize tools/list_changed resyncs with latest-wins coalescing (syncing + pendingResync flags) to prevent concurrent races --- packages/mcp/mcp-client/src/index.ts | 64 ++++++++++++++++++++++------ 1 file changed, 50 insertions(+), 14 deletions(-) diff --git a/packages/mcp/mcp-client/src/index.ts b/packages/mcp/mcp-client/src/index.ts index da3aefc3de..2e797ec199 100644 --- a/packages/mcp/mcp-client/src/index.ts +++ b/packages/mcp/mcp-client/src/index.ts @@ -86,6 +86,16 @@ export const Config = z.union([ // ---- Plugin apply ---- +/** Mutable state shared between the async connect path, notification handler, and disposers. */ +interface PluginState { + /** Current generation of tool disposers (keyed by registered name). */ + disposers: Map void> + /** Whether a syncTools call is currently in-flight. */ + syncing: boolean + /** Whether another tools/list_changed arrived while syncing (coalesce flag). */ + pendingResync: boolean +} + export function apply(ctx: Context, config: Config): void { const transport = createTransport(config) const client = new Client( @@ -93,36 +103,62 @@ export function apply(ctx: Context, config: Config): void { { capabilities: {} }, ) + const state: PluginState = { disposers: new Map(), syncing: false, pendingResync: false } + + const opts = { toolPrefix: config.toolPrefix, toolCallTimeoutMs: config.toolCallTimeoutMs } + + /** Dispose all currently registered tools. */ + function disposeTools(): void { + for (const dispose of state.disposers.values()) dispose() + state.disposers = new Map() + } + + /** Run syncTools with latest-wins coalescing. */ + async function resync(): Promise { + if (state.syncing) { + state.pendingResync = true + return + } + state.syncing = true + try { + state.disposers = await syncTools(client, ctx, opts, state.disposers) + } finally { + state.syncing = false + } + // If another notification arrived while we were syncing, run once more. + if (state.pendingResync) { + state.pendingResync = false + await resync() + } + } + + // When the connection closes (server crash or intentional close), unregister + // all tools so the model no longer sees them in the system prompt. + client.onclose = () => { + disposeTools() + ctx.logger.info('mcp-client: connection closed, tools unregistered') + } + // Connect and set up tools. Errors during connect are logged, not thrown // (the plugin simply has no tools registered). const ready = (async () => { await client.connect(transport) - - let disposers = await syncTools(client, ctx, { - toolPrefix: config.toolPrefix, - toolCallTimeoutMs: config.toolCallTimeoutMs, - }, new Map()) + await resync() client.setNotificationHandler( ToolListChangedNotificationSchema, async () => { ctx.logger.info('mcp-client: tool list changed, re-syncing') - disposers = await syncTools(client, ctx, { - toolPrefix: config.toolPrefix, - toolCallTimeoutMs: config.toolCallTimeoutMs, - }, disposers) + await resync() }, ) - - return disposers })().catch((error: unknown) => { ctx.logger.error(`mcp-client: failed to connect: ${String(error)}`) - return new Map void>() }) + // Fiber disposal: close the client (triggers onclose → tools unregistered). ctx.effect(() => async () => { - const disposers = await ready - for (const dispose of disposers.values()) dispose() + await ready try { await client.close() } catch { /* transport already gone */ } }, 'mcp-client.connection') } From 2efe8ad418e668b7cbe559a65901e4921fc6fabe Mon Sep 17 00:00:00 2001 From: lintianle Date: Wed, 8 Jul 2026 12:18:59 +0800 Subject: [PATCH 03/26] test: cover resync coalescing, onclose, and error path in mcp-client Addresses CI coverage gap: exercises the latest-wins resync coalescing (pendingResync branch), the client.onclose callback, and ensures index.ts is loaded without module mocks for stable v8 coverage across environments. --- packages/mcp/mcp-client/tests/apply.spec.ts | 51 +++++++++++++++++++ .../mcp/mcp-client/tests/mcp-client.spec.ts | 36 ++++++++++++- 2 files changed, 86 insertions(+), 1 deletion(-) diff --git a/packages/mcp/mcp-client/tests/apply.spec.ts b/packages/mcp/mcp-client/tests/apply.spec.ts index 077cb4e3f6..baf1a4985c 100644 --- a/packages/mcp/mcp-client/tests/apply.spec.ts +++ b/packages/mcp/mcp-client/tests/apply.spec.ts @@ -22,6 +22,7 @@ class MockClient { listTools = mockListTools callTool = mockCallTool setNotificationHandler = mockSetNotificationHandler + onclose: (() => void) | null = null } vi.mock('@modelcontextprotocol/sdk/client/index.js', () => ({ @@ -176,4 +177,54 @@ describe('apply (plugin lifecycle)', () => { expect(mockConnect).toHaveBeenCalled() expect(ctx.tools.get('remote')).toBeDefined() }) + + it('coalesces overlapping resync notifications (latest-wins)', async () => { + apply(ctx, stdioConfig) + await new Promise(r => setTimeout(r, 50)) + + // Initial sync is done; notification handler is registered. + const handler = mockSetNotificationHandler.mock.calls[0]![1] as () => Promise + + // Make the NEXT listTools call slow so we can trigger a second notification. + let resolveBlocked!: (v: unknown) => void + mockListTools.mockReturnValueOnce(new Promise((r) => { resolveBlocked = r })) + + // Fire first notification — starts a resync that blocks on listTools. + const firstResync = handler() + + // Fire second notification while the first is in-flight — should coalesce. + const secondResync = handler() + + // Resolve the blocked listTools call. + resolveBlocked({ tools: [{ name: 'mid', inputSchema: { type: 'object' } }], nextCursor: undefined }) + + // Set up the response for the deferred resync that fires after the first completes. + mockListTools.mockResolvedValueOnce({ + tools: [{ name: 'final', inputSchema: { type: 'object' } }], + nextCursor: undefined, + }) + + await firstResync + await secondResync + await new Promise(r => setTimeout(r, 50)) + + // The deferred resync should have run with the latest tool list. + expect(ctx.tools.get('final')).toBeDefined() + }) + + it('unregisters tools when the server connection closes (onclose)', async () => { + apply(ctx, stdioConfig) + await new Promise(r => setTimeout(r, 50)) + + expect(ctx.tools.get('remote')).toBeDefined() + + // Simulate the MCP client's onclose firing (server crashed or closed). + // The apply() sets `client.onclose = () => {...}` on the mock instance. + // mockConnect receives `this` as the client instance. + const clientInstance = mockConnect.mock.contexts[0] as MockClient + expect(clientInstance.onclose).toBeTypeOf('function') + clientInstance.onclose!() + + expect(ctx.tools.get('remote')).toBeUndefined() + }) }) diff --git a/packages/mcp/mcp-client/tests/mcp-client.spec.ts b/packages/mcp/mcp-client/tests/mcp-client.spec.ts index e81369c0f0..5f5a9bf2f4 100644 --- a/packages/mcp/mcp-client/tests/mcp-client.spec.ts +++ b/packages/mcp/mcp-client/tests/mcp-client.spec.ts @@ -5,7 +5,7 @@ import SystemPrompt from '@deepseek-ai/dsh-system-prompt' import ToolRegistry from '@deepseek-ai/dsh-tools' import { syncTools, type ToolBridgeOptions } from '@deepseek-ai/dsh-mcp-client/src/tools.ts' import { createTransport } from '@deepseek-ai/dsh-mcp-client/src/transport.ts' -import type { Config } from '@deepseek-ai/dsh-mcp-client' +import { apply, name, inject, Config } from '@deepseek-ai/dsh-mcp-client/src/index.ts' // ---- Mock MCP Client ---- @@ -516,3 +516,37 @@ describe('tool execution — non-object args fallback', () => { }) }) +describe('plugin module exports', () => { + it('exports name, inject, and Config schema', () => { + expect(name).toBe('mcp-client') + expect(inject).toEqual(['tools']) + expect(Config).toBeDefined() + }) +}) + +describe('apply (error path, no mocks)', () => { + it('gracefully catches when the MCP server is unreachable', async () => { + const ctx = new Context() + await ctx.plugin(SystemPrompt) + await ctx.plugin(ToolRegistry) + + // Call apply with a command that will fail to spawn/connect. + // The .catch() inside apply logs the error and registers no tools. + apply(ctx, { + transport: 'stdio', + command: '___nonexistent_binary_that_will_fail___', + args: [], + env: {}, + cwd: '', + toolPrefix: '', + toolCallTimeoutMs: 1000, + }) + + // Give the async connect + catch chain time to settle. + await new Promise(r => setTimeout(r, 200)) + + // No tools should be registered since connect failed. + expect(ctx.tools.get('anything')).toBeUndefined() + }) +}) + From c65e05cff168e657cb28b365ef7936d0b6cd30c5 Mon Sep 17 00:00:00 2001 From: lintianle Date: Wed, 8 Jul 2026 12:56:04 +0800 Subject: [PATCH 04/26] test: add MCP client e2e tests with real MCP servers Prove the full MCP protocol flow works end-to-end against real servers: - Self-written fixture server: tool discovery, execution, error handling, image placeholder, toolPrefix, and clean disposal - @modelcontextprotocol/server-everything: echo, get-sum, get-tiny-image - @modelcontextprotocol/server-filesystem: write_file + read_file round-trip, list_directory with world-verification All 15 tests keyless and deterministic (no API key needed). --- packages/mcp/mcp-client/package.json | 5 +- .../mcp/mcp-client/tests/fixture-server.ts | 55 +++ .../mcp/mcp-client/tests/mcp-client.e2e.ts | 318 ++++++++++++++++ pnpm-lock.yaml | 341 ++++++++++++++++++ 4 files changed, 718 insertions(+), 1 deletion(-) create mode 100644 packages/mcp/mcp-client/tests/fixture-server.ts create mode 100644 packages/mcp/mcp-client/tests/mcp-client.e2e.ts diff --git a/packages/mcp/mcp-client/package.json b/packages/mcp/mcp-client/package.json index 69626cc606..638777017d 100644 --- a/packages/mcp/mcp-client/package.json +++ b/packages/mcp/mcp-client/package.json @@ -33,6 +33,9 @@ "devDependencies": { "@deepseek-ai/dsh-tools": "workspace:^", "@deepseek-ai/dsh-llm": "workspace:^", - "cordis": "^4.0.0-rc.6" + "@modelcontextprotocol/server-everything": "^2026.7.4", + "@modelcontextprotocol/server-filesystem": "^2026.7.4", + "cordis": "^4.0.0-rc.6", + "zod": "^4.4.3" } } diff --git a/packages/mcp/mcp-client/tests/fixture-server.ts b/packages/mcp/mcp-client/tests/fixture-server.ts new file mode 100644 index 0000000000..6b97c2b59c --- /dev/null +++ b/packages/mcp/mcp-client/tests/fixture-server.ts @@ -0,0 +1,55 @@ +/** + * Minimal MCP server over stdio for e2e testing of the dsh-mcp-client plugin. + * Registers controlled tools with predictable behavior for asserting edge cases. + * + * Run: node --import tsx fixture-server.ts + */ + +import { McpServer } from '@modelcontextprotocol/sdk/server/mcp.js' +import { StdioServerTransport } from '@modelcontextprotocol/sdk/server/stdio.js' +import { z } from 'zod' + +const server = new McpServer( + { name: 'fixture-server', version: '1.0.0' }, + { capabilities: { tools: { listChanged: true } } }, +) + +server.registerTool('add', { + title: 'Add Tool', + description: 'Adds two numbers.', + inputSchema: { a: z.number().describe('First number'), b: z.number().describe('Second number') }, +}, async args => ({ + content: [{ type: 'text', text: String(args.a + args.b) }], +})) + +server.registerTool('greet', { + title: 'Greet Tool', + description: 'Greets a person by name.', + inputSchema: { name: z.string().describe('Name to greet') }, +}, async args => ({ + content: [{ type: 'text', text: `Hello, ${args.name}!` }], +})) + +server.registerTool('fail', { + title: 'Fail Tool', + description: 'Always returns an error.', + inputSchema: {}, +}, async () => ({ + content: [{ type: 'text', text: 'Something went wrong' }], + isError: true, +})) + +server.registerTool('image', { + title: 'Image Tool', + description: 'Returns an image content block.', + inputSchema: {}, +}, async () => ({ + content: [ + { type: 'text', text: 'Here is an image:' }, + { type: 'image', data: 'iVBORw0KGgo=', mimeType: 'image/png' }, + { type: 'text', text: 'End of image.' }, + ], +})) + +const transport = new StdioServerTransport() +await server.connect(transport) diff --git a/packages/mcp/mcp-client/tests/mcp-client.e2e.ts b/packages/mcp/mcp-client/tests/mcp-client.e2e.ts new file mode 100644 index 0000000000..d59ba59d66 --- /dev/null +++ b/packages/mcp/mcp-client/tests/mcp-client.e2e.ts @@ -0,0 +1,318 @@ +/** + * End-to-end tests for dsh-mcp-client. Exercises the REAL MCP protocol over + * stdio transport against: + * 1. A self-written fixture server (controlled edge cases) + * 2. @modelcontextprotocol/server-everything (official integration test server) + * 3. @modelcontextprotocol/server-filesystem (real filesystem operations) + * + * No API key needed — all servers are local/keyless. + */ + +import { mkdtemp, rm, writeFile, readFile } from 'node:fs/promises' +import { tmpdir } from 'node:os' +import { join } from 'node:path' +import { fileURLToPath } from 'node:url' +import { afterAll, beforeAll, describe, expect, it } from 'vitest' +import { Context } from 'cordis' +import SystemPrompt from '@deepseek-ai/dsh-system-prompt' +import ToolRegistry from '@deepseek-ai/dsh-tools' +import { CallId } from '@deepseek-ai/dsh-llm' +import type { Config } from '@deepseek-ai/dsh-mcp-client' + +const tsxLoader = fileURLToPath(import.meta.resolve('tsx')) +const fixtureServerPath = fileURLToPath(new URL('./fixture-server.ts', import.meta.url)) +const repoTsconfig = fileURLToPath(new URL('../../../../tsconfig.json', import.meta.url)) + +// Resolve package-local .bin for pnpm-hoisted MCP server binaries. +const packageDir = fileURLToPath(new URL('..', import.meta.url)) +const localBin = join(packageDir, 'node_modules', '.bin') + +// ---- Helpers ---- + +async function mountRegistry(): Promise { + const ctx = new Context() + await ctx.plugin(SystemPrompt) + await ctx.plugin(ToolRegistry) + return ctx +} + +/** Apply the MCP client plugin and wait for tools to be registered. */ +async function applyAndWait(ctx: Context, config: Config, timeoutMs = 20_000): Promise { + const { apply } = await import('@deepseek-ai/dsh-mcp-client/src/index.ts') + const toolsReady = new Promise((resolve, reject) => { + const timer = setTimeout( + () => { reject(new Error(`applyAndWait timed out after ${timeoutMs}ms — no tools/change event`)) }, + timeoutMs, + ) + ctx.on('tools/change', () => { clearTimeout(timer); resolve() }) + }) + apply(ctx, config) + await toolsReady +} + +let callSeq = 0 +function nextCallId(): CallId { + return CallId(`e2e-${++callSeq}`) +} + +// ---- Fixture server tests ---- + +describe('fixture server — controlled scenarios', () => { + let ctx: Context + + const fixtureConfig: Config = { + transport: 'stdio', + command: process.execPath, + args: ['--import', tsxLoader, fixtureServerPath], + env: { TSX_TSCONFIG_PATH: repoTsconfig }, + cwd: packageDir, + toolPrefix: '', + toolCallTimeoutMs: 15_000, + } + + beforeAll(async () => { + ctx = await mountRegistry() + await applyAndWait(ctx, fixtureConfig) + }, 30_000) + + afterAll(async () => { + if (ctx) await ctx.fiber.dispose() + await new Promise(r => setTimeout(r, 200)) + }) + + it('discovers all fixture tools', () => { + const schemas = ctx.tools.schemas() + const names = schemas.map(s => s.name) + expect(names).toContain('add') + expect(names).toContain('greet') + expect(names).toContain('fail') + expect(names).toContain('image') + }) + + it('executes add(2, 3) → "5"', async () => { + const result = await ctx.tools.execute({ + callId: nextCallId(), name: 'add', arguments: { a: 2, b: 3 }, + }) + expect(result.isError).toBe(false) + expect(result.content[0]).toEqual({ type: 'text', text: '5' }) + }) + + it('executes greet("World") → "Hello, World!"', async () => { + const result = await ctx.tools.execute({ + callId: nextCallId(), name: 'greet', arguments: { name: 'World' }, + }) + expect(result.isError).toBe(false) + expect(result.content[0]).toEqual({ type: 'text', text: 'Hello, World!' }) + }) + + it('executes fail() → isError result', async () => { + const result = await ctx.tools.execute({ + callId: nextCallId(), name: 'fail', arguments: {}, + }) + expect(result.isError).toBe(true) + expect(result.content[0]).toMatchObject({ type: 'text' }) + }) + + it('executes image() → image placeholder', async () => { + const result = await ctx.tools.execute({ + callId: nextCallId(), name: 'image', arguments: {}, + }) + expect(result.isError).toBe(false) + const text = (result.content[0] as { type: string; text: string }).text + expect(text).toContain('Here is an image:') + expect(text).toContain('[image: image/png, content discarded]') + expect(text).toContain('End of image.') + }) +}) + +describe('fixture server — toolPrefix', () => { + let ctx: Context + + beforeAll(async () => { + ctx = await mountRegistry() + await applyAndWait(ctx, { + transport: 'stdio', + command: process.execPath, + args: ['--import', tsxLoader, fixtureServerPath], + env: { TSX_TSCONFIG_PATH: repoTsconfig }, + cwd: packageDir, + toolPrefix: 'fx_', + toolCallTimeoutMs: 15_000, + }) + }, 30_000) + + afterAll(async () => { + if (ctx) await ctx.fiber.dispose() + await new Promise(r => setTimeout(r, 200)) + }) + + it('registers tools with prefix', () => { + expect(ctx.tools.get('fx_add')).toBeDefined() + expect(ctx.tools.get('fx_greet')).toBeDefined() + expect(ctx.tools.get('add')).toBeUndefined() + }) + + it('executes prefixed tool', async () => { + const result = await ctx.tools.execute({ + callId: nextCallId(), name: 'fx_add', arguments: { a: 10, b: 20 }, + }) + expect(result.content[0]).toEqual({ type: 'text', text: '30' }) + }) +}) + +describe('fixture server — disposal', () => { + it('disposes cleanly without error', async () => { + const ctx = await mountRegistry() + await applyAndWait(ctx, { + transport: 'stdio', + command: process.execPath, + args: ['--import', tsxLoader, fixtureServerPath], + env: { TSX_TSCONFIG_PATH: repoTsconfig }, + cwd: packageDir, + toolPrefix: '', + toolCallTimeoutMs: 15_000, + }) + + // Tools are registered before dispose. + expect(ctx.tools.get('add')).toBeDefined() + expect(ctx.tools.schemas().length).toBeGreaterThanOrEqual(4) + + // Dispose should complete without throwing. + await ctx.fiber.dispose() + await new Promise(r => setTimeout(r, 200)) + }, 30_000) +}) + +// ---- @modelcontextprotocol/server-everything ---- + +describe('server-everything — official test server', () => { + let ctx: Context + + const config: Config = { + transport: 'stdio', + command: join(localBin, 'mcp-server-everything'), + args: ['stdio'], + env: {}, + cwd: '', + toolPrefix: '', + toolCallTimeoutMs: 30_000, + } + + beforeAll(async () => { + ctx = await mountRegistry() + await applyAndWait(ctx, config) + }, 60_000) + + afterAll(async () => { + if (ctx) await ctx.fiber.dispose() + await new Promise(r => setTimeout(r, 500)) + }) + + it('discovers tools from server-everything', () => { + const schemas = ctx.tools.schemas() + const names = schemas.map(s => s.name) + expect(names).toContain('echo') + expect(names).toContain('get-sum') + expect(names).toContain('get-tiny-image') + expect(names.length).toBeGreaterThanOrEqual(8) + }) + + it('executes echo({ message: "hello" }) → "Echo: hello"', async () => { + const result = await ctx.tools.execute({ + callId: nextCallId(), name: 'echo', arguments: { message: 'hello' }, + }) + expect(result.isError).toBe(false) + const text = (result.content[0] as { type: string; text: string }).text + expect(text).toBe('Echo: hello') + }) + + it('executes get-sum({ a: 3, b: 7 }) → contains "10"', async () => { + const result = await ctx.tools.execute({ + callId: nextCallId(), name: 'get-sum', arguments: { a: 3, b: 7 }, + }) + expect(result.isError).toBe(false) + const text = (result.content[0] as { type: string; text: string }).text + expect(text).toContain('10') + }) + + it('executes get-tiny-image → image placeholder', async () => { + const result = await ctx.tools.execute({ + callId: nextCallId(), name: 'get-tiny-image', arguments: {}, + }) + expect(result.isError).toBe(false) + const text = (result.content[0] as { type: string; text: string }).text + expect(text).toContain('[image: image/png, content discarded]') + }) +}) + +// ---- @modelcontextprotocol/server-filesystem ---- + +describe('server-filesystem — real filesystem operations', () => { + let ctx: Context + let tempDir: string + + beforeAll(async () => { + tempDir = await mkdtemp(join(tmpdir(), 'mcp-fs-e2e-')) + + ctx = await mountRegistry() + const config: Config = { + transport: 'stdio', + command: join(localBin, 'mcp-server-filesystem'), + args: [tempDir], + env: {}, + cwd: '', + toolPrefix: '', + toolCallTimeoutMs: 30_000, + } + await applyAndWait(ctx, config) + }, 60_000) + + afterAll(async () => { + if (ctx) await ctx.fiber.dispose() + await new Promise(r => setTimeout(r, 500)) + await rm(tempDir, { recursive: true, force: true }) + }) + + it('discovers filesystem tools', () => { + const schemas = ctx.tools.schemas() + const names = schemas.map(s => s.name) + expect(names).toContain('read_file') + expect(names).toContain('write_file') + expect(names).toContain('list_directory') + }) + + it('write_file + read_file round-trip', async () => { + const filePath = join(tempDir, 'test.txt') + const content = 'Hello from MCP e2e test!' + + // Write via MCP tool + const writeResult = await ctx.tools.execute({ + callId: nextCallId(), name: 'write_file', arguments: { path: filePath, content }, + }) + expect(writeResult.isError).toBe(false) + + // Verify file was actually written (world verification) + const onDisk = await readFile(filePath, 'utf8') + expect(onDisk).toBe(content) + + // Read back via MCP tool + const readResult = await ctx.tools.execute({ + callId: nextCallId(), name: 'read_file', arguments: { path: filePath }, + }) + expect(readResult.isError).toBe(false) + const text = (readResult.content[0] as { type: string; text: string }).text + expect(text).toContain(content) + }) + + it('list_directory shows written file', async () => { + // Ensure a file exists + await writeFile(join(tempDir, 'listed.txt'), 'listed') + + const result = await ctx.tools.execute({ + callId: nextCallId(), name: 'list_directory', arguments: { path: tempDir }, + }) + expect(result.isError).toBe(false) + const text = (result.content[0] as { type: string; text: string }).text + expect(text).toContain('listed.txt') + }) +}) diff --git a/pnpm-lock.yaml b/pnpm-lock.yaml index 0c50782144..d12ec0d9c5 100644 --- a/pnpm-lock.yaml +++ b/pnpm-lock.yaml @@ -526,9 +526,18 @@ importers: '@deepseek-ai/dsh-tools': specifier: workspace:^ version: link:../../core/tools + '@modelcontextprotocol/server-everything': + specifier: ^2026.7.4 + version: 2026.7.4 + '@modelcontextprotocol/server-filesystem': + specifier: ^2026.7.4 + version: 2026.7.4(zod@4.4.3) cordis: specifier: ^4.0.0-rc.6 version: 4.0.0-rc.6(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.4) + zod: + specifier: ^4.4.3 + version: 4.4.3 packages/session-persistence/session-persistence: devDependencies: @@ -1723,6 +1732,10 @@ packages: '@iconify/utils@3.1.3': resolution: {integrity: sha512-LPKOXPn/zV+zis1oOfGWogaXVpqUybF3ZS6SCZIsz8vg0ivVp9+fVqyYB7xq0aiST/VhUQYGO1qo6uoYSiEJqw==} + '@isaacs/cliui@8.0.2': + resolution: {integrity: sha512-O8jcjabXaleOG9DQ0+ARXWZBTfnP4WNAqzuiJK7ll44AmxGKv/J2M4TPjxjY3znBCfvBXFzucm1twdyFybFqEA==} + engines: {node: '>=12'} + '@jridgewell/gen-mapping@0.3.13': resolution: {integrity: sha512-2kkt/7niJ6MgEPxF0bYdQ6etZaA+fQvDcLKckhy1yIQOzaoKjBBjSj63/aLVjYE3qhRt5dvM+uUyfCg6UKCBbA==} @@ -1752,6 +1765,14 @@ packages: '@cfworker/json-schema': optional: true + '@modelcontextprotocol/server-everything@2026.7.4': + resolution: {integrity: sha512-ydMW/M6rk9tK23b+U38trsNLHhd5eF+ntiv2Vr+RPMDhbiKY/IKrZU25ukvSXVPUBvy7TxTPWpeV4KcYcXg72w==} + hasBin: true + + '@modelcontextprotocol/server-filesystem@2026.7.4': + resolution: {integrity: sha512-JwEaH4dRRzwcNMwX8WJVCJyXfFxXjFKdgwHxjQhFLhi02kszgyyj611LV9puBLDO1IiDQSCjfKFSPaemegnvwg==} + hasBin: true + '@napi-rs/wasm-runtime@1.1.5': resolution: {integrity: sha512-AWPoBRJ9tsnVhor4sjO7rkni+7p+2IAEFj6cx06UgP10jkQHqay/36uRV/bFkgrh18D9vb4cr8Q0Pthskgzy+Q==} peerDependencies: @@ -1997,6 +2018,10 @@ packages: cpu: [x64] os: [win32] + '@pkgjs/parseargs@0.11.0': + resolution: {integrity: sha512-+1VkjdD0QBLPodGrJUeqarH8VAIvQODIbwh9XpP5Syisf7YoQgsJKPNFoqqLQlu+VQ/tVSshMR6loPMn8U+dPg==} + engines: {node: '>=14'} + '@protobufjs/aspromise@1.1.2': resolution: {integrity: sha512-j+gKExEuLmKwvz3OgROXtrJ2UG2x8Ch2YZUxahh+s1F2HZ+wAceUNLkvy6zKCPVRkU++ZWQrdxsUeQXmcg4uoQ==} @@ -2558,6 +2583,22 @@ packages: ajv@8.20.0: resolution: {integrity: sha512-Thbli+OlOj+iMPYFBVBfJ3OmCAnaSyNn4M1vz9T6Gka5Jt9ba/HIR56joy65tY6kx/FCF5VXNB819Y7/GUrBGA==} + ansi-regex@5.0.1: + resolution: {integrity: sha512-quJQXlTSUGL2LH9SUXo8VwsY4soanhgo6LNSm84E1LBcE8s3O0wpdiRzyR9z/ZZJMlMWv37qOOb9pdJlMUEKFQ==} + engines: {node: '>=8'} + + ansi-regex@6.2.2: + resolution: {integrity: sha512-Bq3SmSpyFHaWjPk8If9yc6svM8c56dB5BAtW4Qbw5jHTwwXXcTLoRMkpDJp6VL0XzlWaCHTXrkFURMYmD0sLqg==} + engines: {node: '>=12'} + + ansi-styles@4.3.0: + resolution: {integrity: sha512-zbB9rCJAT1rbjiVDb2hqKFHNYLxgtk8NURxZ3IZwD3F6NtxbXZQCnnSi1Lkx+IDohdPlFp222wVALIheZJQSEg==} + engines: {node: '>=8'} + + ansi-styles@6.2.3: + resolution: {integrity: sha512-4Dj6M28JB+oAH8kFkTLUo+a2jwOFkuqb3yucU0CANcRRUbxS0cP0nZYCGjcc3BNXwRIsUVmDGgzawme7zvJHvg==} + engines: {node: '>=12'} + ansis@4.3.1: resolution: {integrity: sha512-BJ8/l4R5LRE7hW9WdSuGYrLSHi2ynxeFpDFbH0K/CgNeY/tyhk+vO6TYxXC5r5CpUhNVX310xzPsN/H9lCdfOA==} engines: {node: '>=14'} @@ -2579,6 +2620,9 @@ packages: ast-v8-to-istanbul@1.0.4: resolution: {integrity: sha512-0bC0/4bTSrnwdhU3IsZDwEdojvuPrSg59OYZfKsLRtJZ0u8VBx9DebfqqG8bRdCC0I7vjgxmPi41P0lpkhJHtA==} + balanced-match@1.0.2: + resolution: {integrity: sha512-3oSeUO0TMV67hN1AmbXsK4yaqU7tjiHlbxRDZOpH0KW9+CeX4bRAaX0Anxt0tx2MrpRpWwQaPwIlISEJhYU5Pw==} + balanced-match@4.0.4: resolution: {integrity: sha512-BLrgEcRTwX2o6gGxGOCNyMvGSp35YofuYzw9h1IMTRmKqttAZZVU67bdb9Pr2vUHA8+j3i2tJfjO6C6+4myGTA==} engines: {node: 18 || 20 || >=22} @@ -2602,6 +2646,9 @@ packages: bowser@2.14.1: resolution: {integrity: sha512-tzPjzCxygAKWFOJP011oxFHs57HzIhOEracIgAePE4pqB3LikALKnSzUyU4MGs9/iCEUuHlAJTjTc5M+u7YEGg==} + brace-expansion@2.1.1: + resolution: {integrity: sha512-WR1cURNjuvBLMZBMbqM0UoE+WAfdUcEV1ccD8PVBVOI+Z3ND4+SZbN8RsfT2bMuG1qwz5RFvPukSZm5fF2D5eA==} + brace-expansion@5.0.6: resolution: {integrity: sha512-kLpxurY4Z4r9sgMsyG0Z9uzsBlgiU/EFKhj/h91/8yHu0edo7XuixOIH3VcJ8kkxs6/jPzoI6U9Vj3WqbMQ94g==} engines: {node: 18 || 20 || >=22} @@ -2639,6 +2686,13 @@ packages: resolution: {integrity: sha512-Qgzu8kfBvo+cA4962jnP1KkS6Dop5NS6g7R5LFYJr4b8Ub94PPQXUksCw9PvXoeXPRRddRNC5C1JQUR2SMGtnA==} engines: {node: '>= 14.16.0'} + color-convert@2.0.1: + resolution: {integrity: sha512-RRECPsj7iu/xb5oKYcsFHSppFNnsj/52OVTRKb4zP5onXwVF3zVmmToNcOfGC+CRDpfK/U584fMg38ZHCaElKQ==} + engines: {node: '>=7.0.0'} + + color-name@1.1.4: + resolution: {integrity: sha512-dOy+3AuW3a2wNbZHIuMZpTcgjGuLU/uBL/ubcZF9OXbDo8ff4O8yVp5Bf0efS8uEoYo5q4Fx7dY9OgQGXgAsQA==} + commander@7.2.0: resolution: {integrity: sha512-QrWXB+ZQSVPmIWIhtEO9H+gwHaMGYiF5ChvoJ+K9ZGHG/sVsa6yiesAD1GC/x46sET00Xlwo1u49RVVVzvcSkw==} engines: {node: '>= 10'} @@ -2682,6 +2736,9 @@ packages: '@cordisjs/plugin-loader': optional: true + core-util-is@1.0.3: + resolution: {integrity: sha512-ZQBvi1DcpJ4GDqanjucZ2Hj3wEO5pZDS89BWbkcrvdxksJorwUDDZamX9ldFkp9aw2lmBDLgkObEA4DWNJ9FYQ==} + cors@2.8.6: resolution: {integrity: sha512-tJtZBBHA6vjIAaF6EnIaq6laBBP9aq/Y3ouVJjEfoHbRBcHBAHYcMh/w8LDrk2PvIMMq8gmopa5D4V8RmbrxGw==} engines: {node: '>= 0.10'} @@ -2909,6 +2966,10 @@ packages: devlop@1.1.0: resolution: {integrity: sha512-RWmIqhcFf1lRYBvNmr7qTNuyCt/7/ns2jbpp1+PalgE/rDQcBT0fioSMUpJ93irlUhC5hrg4cYqe6U+0ImW0rA==} + diff@8.0.4: + resolution: {integrity: sha512-DPi0FmjiSU5EvQV0++GFDOJ9ASQUVFh5kD+OzOnYdi7n3Wpm9hWWGfB/O2blfHcMVTL5WkQXSnRiK9makhrcnw==} + engines: {node: '>=0.3.1'} + diff@9.0.0: resolution: {integrity: sha512-svtcdpS8CgJyqAjEQIXdb3OjhFVVYjzGAPO8WGCmRbrml64SPw/jJD4GoE98aR7r25A0XcgrK3F02yw9R/vhQw==} engines: {node: '>=0.3.1'} @@ -2929,12 +2990,21 @@ packages: resolution: {integrity: sha512-KIN/nDJBQRcXw0MLVhZE9iQHmG68qAVIBg9CqmUYjmQIhgij9U5MFvrqkUL5FbtyyzZuOeOt0zdeRe4UY7ct+A==} engines: {node: '>= 0.4'} + eastasianwidth@0.2.0: + resolution: {integrity: sha512-I88TYZWc9XiYHRQ4/3c5rjjfgkjhLyW2luGIheGERbNQ6OY7yTybanSpDXZa8y7VUP9YmDcYa+eyq4ca7iLqWA==} + ecdsa-sig-formatter@1.0.11: resolution: {integrity: sha512-nagl3RYrbNv6kQkeJIpt6NJZy8twLB/2vtz6yN9Z4vRKHN4/QZJIEbqohALSgwKdnksuY3k5Addp5lg8sVoVcQ==} ee-first@1.1.1: resolution: {integrity: sha512-WMwm9LhRUo+WUaRN+vRuETqG89IgZphVSNkdFgeb6sS/E4OrDIN7t48CAewSHXc6C8lefD8KKfr5vY61brQlow==} + emoji-regex@8.0.0: + resolution: {integrity: sha512-MSjYzcWNOA0ewAHpz0MxpYFvwg6yjy1NG3xteoqz644VCo/RPgnr1/GGt+ic3iJTzQ8Eu3TdM14SawnVUmGE6A==} + + emoji-regex@9.2.2: + resolution: {integrity: sha512-L18DaJsXSUk2+42pv8mLs5jJT2hqFkFE4j21wOmgbUqsZ2hL72NsUU785g9RXgo3s0ZNgVl42TiHp3ZtOv/Vyg==} + empathic@2.0.1: resolution: {integrity: sha512-YGRs8knHhKHVShLkFET/rWAU8kmHbOV5LwN938RHI0pljAJ1Gf6SzXsSmRaEzcXTtOOmVqJ5+WtQPL5uigY50Q==} engines: {node: '>=14'} @@ -3121,6 +3191,10 @@ packages: flatted@3.4.2: resolution: {integrity: sha512-PjDse7RzhcPkIJwy5t7KPWQSZ9cAbzQXcafsetQoD7sOJRQlGikNbx7yZp2OotDnJyrDcbyRq3Ttb18iYOqkxA==} + foreground-child@3.3.1: + resolution: {integrity: sha512-gIXjKqtFuWEgzFRJA9WCQeSJLZDjgJUOMCMzxtvFq/37KojM1BFGufqsCy0r4qSQmYLsZYMeyRqzIWOMup03sw==} + engines: {node: '>=14'} + formatly@0.3.0: resolution: {integrity: sha512-9XNj/o4wrRFyhSMJOvsuyMwy8aUfBaZ1VrqHVfohyXf0Sw0e+yfKG+xZaY3arGCOMdwFsqObtzVOc1gU9KiT9w==} engines: {node: '>=18.3.0'} @@ -3173,6 +3247,11 @@ packages: resolution: {integrity: sha512-XxwI8EOhVQgWp6iDL+3b0r86f4d6AX6zSU55HfB4ydCEuXLXc5FcYeOu+nnGftS4TEju/11rt4KJPTMgbfmv4A==} engines: {node: '>=10.13.0'} + glob@10.5.0: + resolution: {integrity: sha512-DfXN8DfhJ7NH3Oe7cFmu3NCu1wKbkReJ8TorzSAFbSKrlNaQSKfIzqYqVY8zlbs2NLBbWpRiU52GX2PbaBVNkg==} + deprecated: Old versions of glob are not supported, and contain widely publicized security vulnerabilities, which have been fixed in the current version. Please update. Support for old versions may be purchased (at exorbitant rates) by contacting i@izs.me + hasBin: true + globrex@0.1.2: resolution: {integrity: sha512-uHJgbwAMwNFf5mLst7IWLNg14x1CkeqglJb/K3doi4dw6q2IvAAmM/Y81kevy83wP+Sst+nutFTYOGg3d1lsxg==} @@ -3245,6 +3324,9 @@ packages: resolution: {integrity: sha512-Hs59xBNfUIunMFgWAbGX5cq6893IbWg4KnrjbYwX3tx0ztorVgTDA6B2sxf8ejHJ4wz8BqGUMYlnzNBer5NvGg==} engines: {node: '>= 4'} + immediate@3.0.6: + resolution: {integrity: sha512-XXOFtyqDjNDAQxVfYxuF7g9Il/IbWmmlQg2MYKOH8ExIT1qg6xc4zyS3HaEEATgs1btfzxq15ciUiY7gjSXRGQ==} + import-meta-resolve@4.2.0: resolution: {integrity: sha512-Iqv2fzaTQN28s/FwZAoFq0ZSs/7hMAHJVX+w8PZl3cY19Pxk6jFFalxQoIfW2826i/fDLXv8IiEZRIT0lDuWcg==} @@ -3278,6 +3360,10 @@ packages: resolution: {integrity: sha512-SbKbANkN603Vi4jEZv49LeVJMn4yGwsbzZworEoyEiutsN3nJYdbO36zfhGJ6QEDpOZIFkDtnq5JRxmvl3jsoQ==} engines: {node: '>=0.10.0'} + is-fullwidth-code-point@3.0.0: + resolution: {integrity: sha512-zymm5+u+sCsSWyD9qNaejV3DFvhCKclKdizYaJUuHA83RLjb7nSuGnddCHGv0hk+KY7BMAlsWeK4Ueg6EV6XQg==} + engines: {node: '>=8'} + is-glob@4.0.3: resolution: {integrity: sha512-xelSayHH36ZgE7ZWhli7pW34hNbNl8Ojv5KVmkJD4hBdD3th8Tfk9vYasLM+mXWOZhFkgZfxhLSnrwRr4elSSg==} engines: {node: '>=0.10.0'} @@ -3288,6 +3374,9 @@ packages: is-promise@4.0.0: resolution: {integrity: sha512-hvpoI6korhJMnej285dSg6nu1+e6uxs7zG3BYAm5byqDsgJNWwxzM6z6iZiAgQR4TJ30JmBTOwqZUw3WlyH3AQ==} + isarray@1.0.0: + resolution: {integrity: sha512-VLghIWNM6ELQzo7zwmcg0NmTVyWKYjvIeM83yjp0wRDTmUnrM678fQbcKBo6n2CJEF0szoG//ytg+TKla89ALQ==} + isexe@2.0.0: resolution: {integrity: sha512-RHxMLp9lnKHGHRng9QFhRCMbYAcVpn69smSGcq3f36xjgVVWThj4qqLbTLlq7Ssj8B+fIQ1EuCEGI2lKsyQeIw==} @@ -3303,6 +3392,9 @@ packages: resolution: {integrity: sha512-HGYWWS/ehqTV3xN10i23tkPkpH46MLCIMFNCaaKNavAXTF1RkqxawEPtnjnGZ6XKSInBKkiOA5BKS+aZiY3AvA==} engines: {node: '>=8'} + jackspeak@3.4.3: + resolution: {integrity: sha512-OGlZQpz2yfahA/Rd1Y8Cd9SIEsqvXkLVoSw/cgwhnhFMDbsQFeZYoJJ7bIZBS9BcamUW96asq/npPWugM+RQBw==} + jiti@2.7.0: resolution: {integrity: sha512-AC/7JofJvZGrrneWNaEnJeOLUx+JlGt7tNa0wZiRPT4MY1wmfKjt2+6O2p2uz2+skll8OZZmJMNqeke7kKbNgQ==} hasBin: true @@ -3356,6 +3448,9 @@ packages: json-stable-stringify-without-jsonify@1.0.1: resolution: {integrity: sha512-Bdboy+l7tA3OGW6FjyFHWkP5LuByj1Tk33Ljyq0axyzdk9//JSi2u3fP1QSmd1KNwq6VOKYGlAu87CisVir6Pw==} + jszip@3.10.1: + resolution: {integrity: sha512-xXDvecyTpGLrqFrvkrUSoxxfJI5AH7U8zxxtVclpsUtMCq4JQ290LY8AW5c7Ggnr/Y/oK+bQMbqK2qmtk3pN4g==} + jwa@2.0.1: resolution: {integrity: sha512-hRF04fqJIP8Abbkq5NKGN0Bbr3JxlQ+qhZufXVr0DvujKy93ZCbXZMHDL4EOtodSbCWxOqR8MS1tXA5hwqCXDg==} @@ -3441,6 +3536,9 @@ packages: resolution: {integrity: sha512-+bT2uH4E5LGE7h/n3evcS/sQlJXCpIp6ym8OWJ5eV6+67Dsql/LaaT7qJBAt2rzfoa/5QBGBhxDix1dMt2kQKQ==} engines: {node: '>= 0.8.0'} + lie@3.3.0: + resolution: {integrity: sha512-UaiMJzeWRlEujzAuw5LokY1L5ecNQYZKfmyZ9L7wDHb/p5etKaxXhohBcrw0EYby+G/NA52vRSN4N39dxHAIwQ==} + lightningcss-android-arm64@1.32.0: resolution: {integrity: sha512-YK7/ClTt4kAK0vo6w3X+Pnm0D2cf2vPHbhOXdoNti1Ga0al1P4TBZhwjATvjNwLEBCnKvjJc2jQgHXH0NEwlAg==} engines: {node: '>= 12.0.0'} @@ -3528,6 +3626,9 @@ packages: longest-streak@3.1.0: resolution: {integrity: sha512-9Ri+o0JYgehTaVBBDoMqIl8GXtbWg711O3srftcHhZ0dqnETqLaoIK0x17fUw9rFSlK/0NlsKe0Ahhyl5pXE2g==} + lru-cache@10.4.3: + resolution: {integrity: sha512-JNAzZcXrCt42VGLuYz0zfAzDfAvJWW6AfYlDBQyDV5DClI2m5sAmK+OIO7s59XfsRsWHp02jAJrRadPRGTt6SQ==} + lru-cache@11.5.1: resolution: {integrity: sha512-RPimw/7aMdv2oqRrxKwvZXcPfwBrn/JZ2xYcY9Hus/6LaS3VOAKVWKWgNLCFSiOm1ESXinjsDlidVU7JlnCN2A==} engines: {node: 20 || >=22} @@ -3697,6 +3798,14 @@ packages: resolution: {integrity: sha512-MULkVLfKGYDFYejP07QOurDLLQpcjk7Fw+7jXS2R2czRQzR56yHRveU5NDJEOviH+hETZKSkIk5c+T23GjFUMg==} engines: {node: 18 || 20 || >=22} + minimatch@9.0.9: + resolution: {integrity: sha512-OBwBN9AL4dqmETlpS2zasx+vTeWclWzkblfZk7KTA5j3jeOONz/tRCnZomUyvNg83wL5Zv9Ss6HMJXAgL8R2Yg==} + engines: {node: '>=16 || 14 >=14.17'} + + minipass@7.1.3: + resolution: {integrity: sha512-tEBHqDnIoM/1rXME1zgka9g6Q2lcoCkxHLuc7ODJ5BxbP5d4c2Z5cGgtXAku59200Cx7diuHTOYfSBD8n6mm8A==} + engines: {node: '>=16 || 14 >=14.17'} + mri@1.2.0: resolution: {integrity: sha512-tzzskb3bG8LvYGFF/mDTpq3jpI6Q9wc3LEmBaghu+DdCssd1FakN7Bc0hVNmEyGq1bq3RgfkCb3cmQLpNPOroA==} engines: {node: '>=4'} @@ -3779,9 +3888,15 @@ packages: resolution: {integrity: sha512-312Id396EbJdvRONlngUx0NydfrIQ5lsYu0znKVUzVvArzEIt08V1qhtyESbGVd1FGX7UKtiFp5uwKZdM8wIuQ==} engines: {node: '>=8'} + package-json-from-dist@1.0.1: + resolution: {integrity: sha512-UEZIS3/by4OC8vL3P2dTXRETpebLI2NiI5vIrjaD/5UtrkFX/tNbwjTSRAGC/+7CAo2pIcBaRgWmcBBHcsaCIw==} + package-manager-detector@1.6.0: resolution: {integrity: sha512-61A5ThoTiDG/C8s8UMZwSorAGwMJ0ERVGj2OjoW5pAalsNOg15+iQiPzrLJ4jhZ1HJzmC2PIHT2oEiH3R5fzNA==} + pako@1.0.11: + resolution: {integrity: sha512-4hLB8Py4zZce5s4yd9XzopqwVv/yGNhV1Bl8NTmCq1763HeK2+EwVTv+leGeL13Dnh2wfbqowVPXCIO0z4taYw==} + parse5@8.0.1: resolution: {integrity: sha512-z1e/HMG90obSGeidlli3hj7cbocou0/wa5HacvI3ASx34PecNjNQeaHNo5WIZpWofN9kgkqV1q5YvXe3F0FoPw==} @@ -3807,6 +3922,10 @@ packages: resolution: {integrity: sha512-ojmeN0qd+y0jszEtoY48r0Peq5dwMEkIlCOu6Q5f41lfkswXuKtYrhgoTpLnyIcHm24Uhqx+5Tqm2InSwLhE6Q==} engines: {node: '>=8'} + path-scurry@1.11.1: + resolution: {integrity: sha512-Xa4Nw17FS9ApQFJ9umLiJS4orGjm7ZzwUrwamcGQuHSzDyth9boKDaycYdDcZDuqYATXw4HFXgaqWTctW/v1HA==} + engines: {node: '>=16 || 14 >=14.18'} + path-to-regexp@8.4.2: resolution: {integrity: sha512-qRcuIdP69NPm4qbACK+aDogI5CBDMi1jKe0ry5rSQJz8JVLsC7jV8XpiJjGRLLol3N+R5ihGYcrPLTno6pAdBA==} @@ -3838,6 +3957,9 @@ packages: resolution: {integrity: sha512-vkcDPrRZo1QZLbn5RLGPpg/WmIQ65qoWWhcGKf/b5eplkkarX0m9z8ppCat4mlOqUsWpyNuYgO3VRyrYHSzX5g==} engines: {node: '>= 0.8.0'} + process-nextick-args@2.0.1: + resolution: {integrity: sha512-3ouUOpQhtgrbOa17J7+uxOTpITYWaGP7/AhoR3+A+/1e9skrzelGi/dXzEYyvbxubEF6Wn2ypscTKiKJFFn1ag==} + protobufjs@7.6.4: resolution: {integrity: sha512-RJJPTTpvFfHcWLkIa2JFWK4XvtSzS0yEWDmunqHXli1h3JlkbcQZXDZdcWxv+JK3Xsl5/UFDPZ0iGm7DAengYw==} engines: {node: '>=12.0.0'} @@ -3873,6 +3995,9 @@ packages: resolution: {integrity: sha512-K5zQjDllxWkf7Z5xJdV0/B0WTNqx6vxG70zJE4N0kBs4LovmEYWJzQGxC9bS9RAKu3bgM40lrd5zoLJ12MQ5BA==} engines: {node: '>= 0.10'} + readable-stream@2.3.8: + resolution: {integrity: sha512-8p0AUk4XODgIewSi0l8Epjs+EVnWiK7NoDIEGU0HhE7+ZyY8D1IMY7odu5lRrFXGg71L15KG8QrPmum45RTtdA==} + readdirp@4.1.2: resolution: {integrity: sha512-GDhwkLfywWL2s6vEjyhri+eXmfH6j1L7JE27WhqLeYzoh/A3DBaYGEj2H/HFZCn/kMfim73FXxEJTw06WtxQwg==} engines: {node: '>= 14.18.0'} @@ -3934,6 +4059,9 @@ packages: resolution: {integrity: sha512-xal3CZX1Xlo/k4ApwCFrHVACi9fBqJ7V+mwhBsuf/1IOKbBy098Fex+Wa/5QMubw09pSZ/u8EY8PWgevJsXp1A==} engines: {node: '>=6'} + safe-buffer@5.1.2: + resolution: {integrity: sha512-Gd2UZBJDkXlY7GbJxfsE8/nvKkUEU1G38c1siN6QP6a9PT9MmHB8GnpscSmMJSoF8LOIrt8ud/wPtojys4G6+g==} + safe-buffer@5.2.1: resolution: {integrity: sha512-rp3So07KcdmmKbGvgaNxQSJr7bGVSVk5S9Eq1F+ppbRo70+YeaDxkw5Dd8NPN+GD6bjnYm2VuPuCXmpuYvmCXQ==} @@ -3960,6 +4088,9 @@ packages: resolution: {integrity: sha512-xRXBn0pPqQTVQiC8wyQrKs2MOlX24zQ0POGaj0kultvoOCstBQM5yvOhAVSUwOMjQtTvsPWoNCHfPGwaaQJhTw==} engines: {node: '>= 18'} + setimmediate@1.0.5: + resolution: {integrity: sha512-MATJdZp8sLqDl/68LfQmbP8zKPLQNV6BIZoIgrscFDQ+RsvK/BxeDQOgyxKKoh0y/8h3BqVFnCqQ/gd+reiIXA==} + setprototypeof@1.2.0: resolution: {integrity: sha512-E5LDX7Wrp85Kil5bhZv46j8jOeboKq5JMmYM3gVGdGH8xFpPWXUMsNrlODCrkoxMEeNi/XZIwuRvY4XNwYMJpw==} @@ -3990,6 +4121,10 @@ packages: siginfo@2.0.0: resolution: {integrity: sha512-ybx0WO1/8bSBLEWXZvEd7gMW3Sn3JFlW3TvX1nREbDLRNQNaeNN8WK0meBwPdAaOI7TtRRRJn/Es1zhrrCHu7g==} + signal-exit@4.1.0: + resolution: {integrity: sha512-bzyZ1e88w9O1iNJbKnOlvYTrWPDl46O1bG0D3XInv+9tkPrxrN8jUUTiFlDkkmKWgn1M6CfIA13SuGqOa9Korw==} + engines: {node: '>=14'} + smol-toml@1.6.1: resolution: {integrity: sha512-dWUG8F5sIIARXih1DTaQAX4SsiTXhInKf1buxdY9DIg4ZYPZK5nGM1VRIYmEbDbsHt7USo99xSLFu5Q1IqTmsg==} engines: {node: '>= 18'} @@ -4008,6 +4143,25 @@ packages: std-env@4.1.0: resolution: {integrity: sha512-Rq7ybcX2RuC55r9oaPVEW7/xu3tj8u4GeBYHBWCychFtzMIr86A7e3PPEBPT37sHStKX3+TiX/Fr/ACmJLVlLQ==} + string-width@4.2.3: + resolution: {integrity: sha512-wKyQRQpjJ0sIp62ErSZdGsjMJWsap5oRNihHhu6G7JVO/9jIB6UyevL+tXuOqrng8j/cxKTWyWUwvSTriiZz/g==} + engines: {node: '>=8'} + + string-width@5.1.2: + resolution: {integrity: sha512-HnLOCR3vjcY8beoNLtcjZ5/nxn2afmME6lhrDrebokqMap+XbeW8n9TXpPDOqdGK5qcI3oT0GKTW6wC7EMiVqA==} + engines: {node: '>=12'} + + string_decoder@1.1.1: + resolution: {integrity: sha512-n/ShnvDi6FHbbVfviro+WojiFzv+s8MPMHBczVePfUpDJLwoLT0ht1l4YwBCbi8pJAveEEdnkHyPyTP/mzRfwg==} + + strip-ansi@6.0.1: + resolution: {integrity: sha512-Y38VPSHcqkFrCpFnQ9vuSXmquuv5oXOKpGeT6aGrr3o3Gc9AlVa6JBfUSOCnbxGGZF+/0ooI7KrPuUSztUdU5A==} + engines: {node: '>=8'} + + strip-ansi@7.2.0: + resolution: {integrity: sha512-yDPMNjp4WyfYBkHnjIRLfca1i6KMyGCtsVgoKe/z1+6vukgaENdgGBZt+ZmKPc4gavvEZ5OgHfHdrazhgNyG7w==} + engines: {node: '>=12'} + strip-json-comments@5.0.3: resolution: {integrity: sha512-1tB5mhVo7U+ETBKNf92xT4hrQa3pm0MZ0PQvuDnWgAAGHDsfp4lPSpiS6psrSiet87wyGPh9ft6wmhOMQ0hDiw==} engines: {node: '>=14.16'} @@ -4192,6 +4346,9 @@ packages: uri-js@4.4.1: resolution: {integrity: sha512-7rKUyy33Q1yc98pQ1DAmLtwX109F7TIfWlW1Ydo8Wl1ii1SeHieeh0HHfPeL2fMXK6z0s8ecKs9frCuLJvndBg==} + util-deprecate@1.0.2: + resolution: {integrity: sha512-EPD5q1uXyFxJpCrLnCc1nHnq3gOa6DZBocAIiI2TaSCA7VCJ1UJDMagCzIkXNsUYfD1daK//LTEQ8xiIbrHtcw==} + uuid@14.0.1: resolution: {integrity: sha512-6ZxzVpzDXDa3bJWaHilVayA+BH/1zmxCJoVgvmqJnid/gPoKHxUrS/aC/T6LGQtNHT+XHG9fXPJB4d+IrU30Ew==} hasBin: true @@ -4327,6 +4484,14 @@ packages: resolution: {integrity: sha512-BN22B5eaMMI9UMtjrGd5g5eCYPpCPDUy0FJXbYsaT5zYxjFOckS53SQDE3pWkVoWpHXVb3BrYcEN4Twa55B5cA==} engines: {node: '>=0.10.0'} + wrap-ansi@7.0.0: + resolution: {integrity: sha512-YVGIj2kamLSTxw6NsZjoBxfSwsn0ycdesmc4p+Q21c5zPuZ1pl+NfxVdxPtdHvmNVOQ6XSYG4AUtyt/Fi7D16Q==} + engines: {node: '>=10'} + + wrap-ansi@8.1.0: + resolution: {integrity: sha512-si7QWI6zUMq56bESFvagtmzMdGOtoxfR+Sez11Mobfc7tm+VkUckk9bW2UeffTGVUbOksxmSw0AA2gs8g71NCQ==} + engines: {node: '>=12'} + wrappy@1.0.2: resolution: {integrity: sha512-l4Sp/DRseor9wL6EvV2+TuQn63dMkPjZ/sp9XkghTEbV9KlPS1xUsZ3u7/IQO4wxtcFB4bgpQPRcR3QCvezPcQ==} @@ -4930,6 +5095,15 @@ snapshots: '@iconify/types': 2.0.0 import-meta-resolve: 4.2.0 + '@isaacs/cliui@8.0.2': + dependencies: + string-width: 5.1.2 + string-width-cjs: string-width@4.2.3 + strip-ansi: 7.2.0 + strip-ansi-cjs: strip-ansi@6.0.1 + wrap-ansi: 8.1.0 + wrap-ansi-cjs: wrap-ansi@7.0.0 + '@jridgewell/gen-mapping@0.3.13': dependencies: '@jridgewell/sourcemap-codec': 1.5.5 @@ -4979,6 +5153,28 @@ snapshots: transitivePeerDependencies: - supports-color + '@modelcontextprotocol/server-everything@2026.7.4': + dependencies: + '@modelcontextprotocol/sdk': 1.29.0(zod@4.4.3) + cors: 2.8.6 + express: 5.2.1 + jszip: 3.10.1 + zod: 4.4.3 + transitivePeerDependencies: + - '@cfworker/json-schema' + - supports-color + + '@modelcontextprotocol/server-filesystem@2026.7.4(zod@4.4.3)': + dependencies: + '@modelcontextprotocol/sdk': 1.29.0(zod@4.4.3) + diff: 8.0.4 + glob: 10.5.0 + minimatch: 10.2.5 + transitivePeerDependencies: + - '@cfworker/json-schema' + - supports-color + - zod + '@napi-rs/wasm-runtime@1.1.5(@emnapi/core@1.10.0)(@emnapi/runtime@1.10.0)': dependencies: '@emnapi/core': 1.10.0 @@ -5124,6 +5320,9 @@ snapshots: '@oxc-resolver/binding-win32-x64-msvc@11.20.0': optional: true + '@pkgjs/parseargs@0.11.0': + optional: true + '@protobufjs/aspromise@1.1.2': {} '@protobufjs/base64@1.1.2': {} @@ -5683,6 +5882,16 @@ snapshots: json-schema-traverse: 1.0.0 require-from-string: 2.0.2 + ansi-regex@5.0.1: {} + + ansi-regex@6.2.2: {} + + ansi-styles@4.3.0: + dependencies: + color-convert: 2.0.1 + + ansi-styles@6.2.3: {} + ansis@4.3.1: {} anynum@1.0.0: {} @@ -5703,6 +5912,8 @@ snapshots: estree-walker: 3.0.3 js-tokens: 10.0.0 + balanced-match@1.0.2: {} + balanced-match@4.0.4: {} base64-js@1.5.1: {} @@ -5731,6 +5942,10 @@ snapshots: bowser@2.14.1: {} + brace-expansion@2.1.1: + dependencies: + balanced-match: 1.0.2 + brace-expansion@5.0.6: dependencies: balanced-match: 4.0.4 @@ -5761,6 +5976,12 @@ snapshots: dependencies: readdirp: 4.1.2 + color-convert@2.0.1: + dependencies: + color-name: 1.1.4 + + color-name@1.1.4: {} + commander@7.2.0: {} commander@8.3.0: {} @@ -5793,6 +6014,8 @@ snapshots: '@cordisjs/plugin-include': link:vendor/include '@cordisjs/plugin-loader': link:vendor/loader + core-util-is@1.0.3: {} + cors@2.8.6: dependencies: object-assign: 4.1.1 @@ -6042,6 +6265,8 @@ snapshots: dependencies: dequal: 2.0.3 + diff@8.0.4: {} + diff@9.0.0: {} dompurify@3.4.11: @@ -6058,12 +6283,18 @@ snapshots: es-errors: 1.3.0 gopd: 1.2.0 + eastasianwidth@0.2.0: {} + ecdsa-sig-formatter@1.0.11: dependencies: safe-buffer: 5.2.1 ee-first@1.1.1: {} + emoji-regex@8.0.0: {} + + emoji-regex@9.2.2: {} + empathic@2.0.1: {} encodeurl@2.0.0: {} @@ -6309,6 +6540,11 @@ snapshots: flatted@3.4.2: {} + foreground-child@3.3.1: + dependencies: + cross-spawn: 7.0.6 + signal-exit: 4.1.0 + formatly@0.3.0: dependencies: fd-package-json: 2.0.0 @@ -6372,6 +6608,15 @@ snapshots: dependencies: is-glob: 4.0.3 + glob@10.5.0: + dependencies: + foreground-child: 3.3.1 + jackspeak: 3.4.3 + minimatch: 9.0.9 + minipass: 7.1.3 + package-json-from-dist: 1.0.1 + path-scurry: 1.11.1 + globrex@0.1.2: {} google-auth-library@10.7.0: @@ -6445,6 +6690,8 @@ snapshots: ignore@7.0.5: {} + immediate@3.0.6: {} + import-meta-resolve@4.2.0: {} import-without-cache@0.4.0: {} @@ -6463,6 +6710,8 @@ snapshots: is-extglob@2.1.1: {} + is-fullwidth-code-point@3.0.0: {} + is-glob@4.0.3: dependencies: is-extglob: 2.1.1 @@ -6471,6 +6720,8 @@ snapshots: is-promise@4.0.0: {} + isarray@1.0.0: {} + isexe@2.0.0: {} istanbul-lib-coverage@3.2.2: {} @@ -6486,6 +6737,12 @@ snapshots: html-escaper: 2.0.2 istanbul-lib-report: 3.0.1 + jackspeak@3.4.3: + dependencies: + '@isaacs/cliui': 8.0.2 + optionalDependencies: + '@pkgjs/parseargs': 0.11.0 + jiti@2.7.0: {} jose@6.2.3: {} @@ -6545,6 +6802,13 @@ snapshots: json-stable-stringify-without-jsonify@1.0.1: {} + jszip@3.10.1: + dependencies: + lie: 3.3.0 + pako: 1.0.11 + readable-stream: 2.3.8 + setimmediate: 1.0.5 + jwa@2.0.1: dependencies: buffer-equal-constant-time: 1.0.1 @@ -6634,6 +6898,10 @@ snapshots: prelude-ls: 1.2.1 type-check: 0.4.0 + lie@3.3.0: + dependencies: + immediate: 3.0.6 + lightningcss-android-arm64@1.32.0: optional: true @@ -6693,6 +6961,8 @@ snapshots: longest-streak@3.1.0: {} + lru-cache@10.4.3: {} + lru-cache@11.5.1: {} magic-string@0.30.21: @@ -7048,6 +7318,12 @@ snapshots: dependencies: brace-expansion: 5.0.6 + minimatch@9.0.9: + dependencies: + brace-expansion: 2.1.1 + + minipass@7.1.3: {} + mri@1.2.0: {} ms@2.1.3: {} @@ -7154,8 +7430,12 @@ snapshots: '@types/retry': 0.12.0 retry: 0.13.1 + package-json-from-dist@1.0.1: {} + package-manager-detector@1.6.0: {} + pako@1.0.11: {} + parse5@8.0.1: dependencies: entities: 8.0.0 @@ -7172,6 +7452,11 @@ snapshots: path-key@3.1.1: {} + path-scurry@1.11.1: + dependencies: + lru-cache: 10.4.3 + minipass: 7.1.3 + path-to-regexp@8.4.2: {} pathe@2.0.3: {} @@ -7197,6 +7482,8 @@ snapshots: prelude-ls@1.2.1: {} + process-nextick-args@2.0.1: {} + protobufjs@7.6.4: dependencies: '@protobufjs/aspromise': 1.1.2 @@ -7243,6 +7530,16 @@ snapshots: iconv-lite: 0.7.3 unpipe: 1.0.0 + readable-stream@2.3.8: + dependencies: + core-util-is: 1.0.3 + inherits: 2.0.4 + isarray: 1.0.0 + process-nextick-args: 2.0.1 + safe-buffer: 5.1.2 + string_decoder: 1.1.1 + util-deprecate: 1.0.2 + readdirp@4.1.2: {} require-from-string@2.0.2: {} @@ -7334,6 +7631,8 @@ snapshots: dependencies: mri: 1.2.0 + safe-buffer@5.1.2: {} + safe-buffer@5.2.1: {} safer-buffer@2.1.2: {} @@ -7374,6 +7673,8 @@ snapshots: transitivePeerDependencies: - supports-color + setimmediate@1.0.5: {} + setprototypeof@1.2.0: {} shebang-command@2.0.0: @@ -7412,6 +7713,8 @@ snapshots: siginfo@2.0.0: {} + signal-exit@4.1.0: {} + smol-toml@1.6.1: {} source-map-js@1.2.1: {} @@ -7422,6 +7725,30 @@ snapshots: std-env@4.1.0: {} + string-width@4.2.3: + dependencies: + emoji-regex: 8.0.0 + is-fullwidth-code-point: 3.0.0 + strip-ansi: 6.0.1 + + string-width@5.1.2: + dependencies: + eastasianwidth: 0.2.0 + emoji-regex: 9.2.2 + strip-ansi: 7.2.0 + + string_decoder@1.1.1: + dependencies: + safe-buffer: 5.1.2 + + strip-ansi@6.0.1: + dependencies: + ansi-regex: 5.0.1 + + strip-ansi@7.2.0: + dependencies: + ansi-regex: 6.2.2 + strip-json-comments@5.0.3: {} strnum@2.4.0: @@ -7577,6 +7904,8 @@ snapshots: dependencies: punycode: 2.3.1 + util-deprecate@1.0.2: {} + uuid@14.0.1: {} vary@1.1.2: {} @@ -7710,6 +8039,18 @@ snapshots: word-wrap@1.2.5: {} + wrap-ansi@7.0.0: + dependencies: + ansi-styles: 4.3.0 + string-width: 4.2.3 + strip-ansi: 6.0.1 + + wrap-ansi@8.1.0: + dependencies: + ansi-styles: 6.2.3 + string-width: 5.1.2 + strip-ansi: 7.2.0 + wrappy@1.0.2: {} ws@8.21.0: {} From fdd7d1a91cb8430435f05f7e104a2779b99ab41d Mon Sep 17 00:00:00 2001 From: lintianle Date: Wed, 8 Jul 2026 13:01:00 +0800 Subject: [PATCH 05/26] chore: register mcp-client e2e entries in knip config Add the mcp-client workspace override so knip recognises the e2e test file, fixture-server entry, and the bin-only devDeps (server-everything, server-filesystem) that are invoked at runtime rather than imported. --- knip.json | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/knip.json b/knip.json index 8c0f71f3af..61ac9b07af 100644 --- a/knip.json +++ b/knip.json @@ -69,6 +69,11 @@ "packages/fs/tool-fs": { "entry": ["tests/**/*.spec.ts", "tests/**/*.e2e.ts"], "project": ["src/**/*.ts", "tests/**/*.ts"] + }, + "packages/mcp/mcp-client": { + "entry": ["tests/**/*.spec.ts", "tests/**/*.e2e.ts", "tests/fixture-server.ts"], + "project": ["src/**/*.ts", "tests/**/*.ts"], + "ignoreDependencies": ["@modelcontextprotocol/server-everything", "@modelcontextprotocol/server-filesystem"] } } } From 0c8f2f7dafcf8d5919ea732fda65ebb219531a92 Mon Sep 17 00:00:00 2001 From: lintianle Date: Wed, 8 Jul 2026 13:23:46 +0800 Subject: [PATCH 06/26] fix: prevent partial tool leaks and non-blocking dispose MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit - syncTools: on paginated listTools failure, unregister any tools already registered in the current sync before rethrowing (prevents orphans) - Effect disposer: call client.close() directly without awaiting startup completion — aborts a hanging connect promptly on HMR/dispose --- packages/mcp/mcp-client/src/index.ts | 12 ++++--- packages/mcp/mcp-client/src/tools.ts | 47 ++++++++++++++++------------ 2 files changed, 34 insertions(+), 25 deletions(-) diff --git a/packages/mcp/mcp-client/src/index.ts b/packages/mcp/mcp-client/src/index.ts index 2e797ec199..a1f0fb1232 100644 --- a/packages/mcp/mcp-client/src/index.ts +++ b/packages/mcp/mcp-client/src/index.ts @@ -140,8 +140,9 @@ export function apply(ctx: Context, config: Config): void { } // Connect and set up tools. Errors during connect are logged, not thrown - // (the plugin simply has no tools registered). - const ready = (async () => { + // (the plugin simply has no tools registered). The IIFE is fire-and-forget; + // disposal closes the client directly without waiting for startup. + void (async () => { await client.connect(transport) await resync() @@ -156,9 +157,10 @@ export function apply(ctx: Context, config: Config): void { ctx.logger.error(`mcp-client: failed to connect: ${String(error)}`) }) - // Fiber disposal: close the client (triggers onclose → tools unregistered). + // Fiber disposal: close the client immediately (triggers onclose → tools + // unregistered). No `await ready` — if connect is still pending, close aborts + // it promptly rather than blocking until the SDK request times out. ctx.effect(() => async () => { - await ready - try { await client.close() } catch { /* transport already gone */ } + try { await client.close() } catch { /* transport already gone or never connected */ } }, 'mcp-client.connection') } diff --git a/packages/mcp/mcp-client/src/tools.ts b/packages/mcp/mcp-client/src/tools.ts index c3a35ccfba..8076a4a764 100644 --- a/packages/mcp/mcp-client/src/tools.ts +++ b/packages/mcp/mcp-client/src/tools.ts @@ -43,27 +43,34 @@ export async function syncTools( const disposers: ToolDisposers = new Map() - let cursor: string | undefined - do { - const response = await client.listTools(cursor ? { cursor } : undefined) - for (const tool of response.tools) { - const registeredName = opts.toolPrefix + tool.name - const definition: ToolDefinition = { - name: registeredName, - description: tool.description ?? '', - parameters: tool.inputSchema, - execute: createExecutor(client, tool.name, opts), + try { + let cursor: string | undefined + do { + const response = await client.listTools(cursor ? { cursor } : undefined) + for (const tool of response.tools) { + const registeredName = opts.toolPrefix + tool.name + const definition: ToolDefinition = { + name: registeredName, + description: tool.description ?? '', + parameters: tool.inputSchema, + execute: createExecutor(client, tool.name, opts), + } + try { + const dispose = ctx.tools.register(definition) + disposers.set(registeredName, dispose) + } catch { + // Name conflict — another tool with this name is already registered. + ctx.logger.warn(`mcp-client: skipping tool "${registeredName}" (name conflict)`) + } } - try { - const dispose = ctx.tools.register(definition) - disposers.set(registeredName, dispose) - } catch { - // Name conflict — another tool with this name is already registered. - ctx.logger.warn(`mcp-client: skipping tool "${registeredName}" (name conflict)`) - } - } - cursor = response.nextCursor - } while (cursor) + cursor = response.nextCursor + } while (cursor) + } catch (error: unknown) { + // Partial failure (e.g. a later page of listTools failed): unregister any + // tools already registered in this sync to avoid orphaning them. + for (const dispose of disposers.values()) dispose() + throw error + } return disposers } From 418b259a11843576fa97f6f7342fcfc489116eac Mon Sep 17 00:00:00 2001 From: lintianle Date: Wed, 8 Jul 2026 13:23:46 +0800 Subject: [PATCH 07/26] fix: prevent partial tool leaks and non-blocking dispose MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit - syncTools: on paginated listTools failure, unregister any tools already registered in the current sync before rethrowing (prevents orphans) - Effect disposer: call client.close() directly without awaiting startup completion — aborts a hanging connect promptly on HMR/dispose --- packages/mcp/mcp-client/tests/mcp-client.spec.ts | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/packages/mcp/mcp-client/tests/mcp-client.spec.ts b/packages/mcp/mcp-client/tests/mcp-client.spec.ts index 5f5a9bf2f4..253cf49d87 100644 --- a/packages/mcp/mcp-client/tests/mcp-client.spec.ts +++ b/packages/mcp/mcp-client/tests/mcp-client.spec.ts @@ -119,6 +119,18 @@ describe('syncTools', () => { expect(secondDisposers.size).toBe(1) }) + it('cleans up already-registered tools when a later page fails', async () => { + const client = createMockClient([]) + client.listTools + .mockResolvedValueOnce({ tools: [{ name: 'survives_not', inputSchema: { type: 'object' } }], nextCursor: 'cursor1' }) + .mockRejectedValueOnce(new Error('page 2 network error')) + + await expect(syncTools(client as never, ctx, defaultOpts, new Map())).rejects.toThrow('page 2 network error') + + // The tool from page 1 was registered then cleaned up on failure. + expect(ctx.tools.get('survives_not')).toBeUndefined() + }) + it('drains paginated listTools responses', async () => { const client = createMockClient([]) client.listTools From a1a78ae30ae73758d57fddfe61803f4302b9eaba Mon Sep 17 00:00:00 2001 From: lintianle Date: Wed, 8 Jul 2026 13:23:46 +0800 Subject: [PATCH 08/26] fix: prevent partial tool leaks and non-blocking dispose MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit - syncTools: on paginated listTools failure, unregister any tools already registered in the current sync before rethrowing (prevents orphans) - Effect disposer: call client.close() directly without awaiting startup completion — aborts a hanging connect promptly on HMR/dispose --- packages/mcp/mcp-client/src/index.ts | 4 +- packages/mcp/mcp-client/src/tools.ts | 92 +++++++++++-------- .../mcp/mcp-client/tests/mcp-client.spec.ts | 30 +++++- 3 files changed, 84 insertions(+), 42 deletions(-) diff --git a/packages/mcp/mcp-client/src/index.ts b/packages/mcp/mcp-client/src/index.ts index a1f0fb1232..c0800b7618 100644 --- a/packages/mcp/mcp-client/src/index.ts +++ b/packages/mcp/mcp-client/src/index.ts @@ -73,14 +73,14 @@ export const Config = z.union([ env: z.dict(String).default({}), cwd: z.string().default(''), toolPrefix: z.string().default(''), - toolCallTimeoutMs: z.number().default(DEFAULT_TOOL_CALL_TIMEOUT_MS), + toolCallTimeoutMs: z.natural().min(1).default(DEFAULT_TOOL_CALL_TIMEOUT_MS), }), z.object({ transport: z.const('streamable-http'), url: z.string().required(), headers: z.dict(String).default({}), toolPrefix: z.string().default(''), - toolCallTimeoutMs: z.number().default(DEFAULT_TOOL_CALL_TIMEOUT_MS), + toolCallTimeoutMs: z.natural().min(1).default(DEFAULT_TOOL_CALL_TIMEOUT_MS), }), ]) as unknown as z diff --git a/packages/mcp/mcp-client/src/tools.ts b/packages/mcp/mcp-client/src/tools.ts index 8076a4a764..e4e9691d88 100644 --- a/packages/mcp/mcp-client/src/tools.ts +++ b/packages/mcp/mcp-client/src/tools.ts @@ -18,19 +18,24 @@ export interface ToolBridgeOptions { /** State for one sync generation: the current set of disposers keyed by tool name. */ type ToolDisposers = Map void> +/** A tool fetched from the MCP server, pending registration. */ +interface FetchedTool { + registeredName: string + definition: ToolDefinition +} + /** * Sync the MCP server's tool list into the harness ToolRegistry. * - * - Calls `client.listTools()` (paginated: drains all pages). - * - Registers each tool as a raw `ToolDefinition`. - * - On name conflict: logs a warning and skips that tool. - * - Returns a disposer map; call each value to unregister. + * Two-phase approach: fetch all pages first (no side effects), then dispose old + * tools and register new ones. If fetching fails, the previous generation stays + * intact — no tools are lost on a transient listTools failure. * * @param client - Connected MCP Client instance used to list and call tools. * @param ctx - Cordis context providing the `tools` service for registration. * @param opts - Bridge options: tool name prefix and per-call timeout. - * @param previous - Disposer map from a prior sync generation; all entries are - * disposed before re-registering. + * @param previous - Disposer map from a prior sync generation; disposed only + * after all pages are successfully fetched. * @returns A map of registered tool names to their unregister disposers. */ export async function syncTools( @@ -39,37 +44,38 @@ export async function syncTools( opts: ToolBridgeOptions, previous: ToolDisposers, ): Promise { - for (const dispose of previous.values()) dispose() - - const disposers: ToolDisposers = new Map() - - try { - let cursor: string | undefined - do { - const response = await client.listTools(cursor ? { cursor } : undefined) - for (const tool of response.tools) { - const registeredName = opts.toolPrefix + tool.name - const definition: ToolDefinition = { + // Phase 1: fetch all tools (no mutations). + const fetched: FetchedTool[] = [] + let cursor: string | undefined + do { + const response = await client.listTools(cursor ? { cursor } : undefined) + for (const tool of response.tools) { + const registeredName = opts.toolPrefix + tool.name + fetched.push({ + registeredName, + definition: { name: registeredName, description: tool.description ?? '', parameters: tool.inputSchema, execute: createExecutor(client, tool.name, opts), - } - try { - const dispose = ctx.tools.register(definition) - disposers.set(registeredName, dispose) - } catch { - // Name conflict — another tool with this name is already registered. - ctx.logger.warn(`mcp-client: skipping tool "${registeredName}" (name conflict)`) - } - } - cursor = response.nextCursor - } while (cursor) - } catch (error: unknown) { - // Partial failure (e.g. a later page of listTools failed): unregister any - // tools already registered in this sync to avoid orphaning them. - for (const dispose of disposers.values()) dispose() - throw error + }, + }) + } + cursor = response.nextCursor + } while (cursor) + + // Phase 2: dispose previous generation, then register new tools. + // If we reach here, all pages were fetched successfully. + for (const dispose of previous.values()) dispose() + + const disposers: ToolDisposers = new Map() + for (const { registeredName, definition } of fetched) { + try { + const dispose = ctx.tools.register(definition) + disposers.set(registeredName, dispose) + } catch { + ctx.logger.warn(`mcp-client: skipping tool "${registeredName}" (name conflict)`) + } } return disposers @@ -129,14 +135,21 @@ function createExecutor( // with optional fallbacks). // eslint-disable-next-line @typescript-eslint/no-unsafe-assignment const content: McpContentBlock[] = result.content - const text = extractText(content, mcpToolName) + let text = extractText(content, mcpToolName) + + // MCP tools with outputSchema may return structuredContent with an empty + // content array. Surface the structured payload as JSON so the model sees + // the actual result. + if (!text && 'structuredContent' in result && result.structuredContent != null) { + text = JSON.stringify(result.structuredContent) + } // MCP isError → throw so ToolRegistry produces an isError result for the model. if ('isError' in result && result.isError === true) { - throw new Error(text) + throw new Error(text || 'MCP tool error') } - return [{ type: 'text', text }] + return [{ type: 'text', text: text || `(${mcpToolName} returned no content)` }] } } @@ -147,8 +160,11 @@ function createExecutor( * * Defensive: fields that the MCP spec declares required (mimeType, text) are * guarded with fallbacks because this is a network trust boundary. + * + * Returns empty string when no text parts were extracted (caller decides + * fallback — e.g. structuredContent). */ -function extractText(mcpContent: McpContentBlock[], toolName: string): string { +function extractText(mcpContent: McpContentBlock[], _toolName: string): string { const parts: string[] = [] for (const block of mcpContent) { @@ -171,5 +187,5 @@ function extractText(mcpContent: McpContentBlock[], toolName: string): string { } } - return parts.join('\n') || `(${toolName} returned no text content)` + return parts.join('\n') } diff --git a/packages/mcp/mcp-client/tests/mcp-client.spec.ts b/packages/mcp/mcp-client/tests/mcp-client.spec.ts index 253cf49d87..8d17f3c64d 100644 --- a/packages/mcp/mcp-client/tests/mcp-client.spec.ts +++ b/packages/mcp/mcp-client/tests/mcp-client.spec.ts @@ -326,7 +326,7 @@ describe('tool execution edge cases', () => { await syncTools(client as never, ctx, defaultOpts, new Map()) const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'notext', arguments: {} }) - expect(result.content[0]).toEqual({ type: 'text', text: '(notext returned no text content)' }) + expect(result.content[0]).toEqual({ type: 'text', text: '(notext returned no content)' }) }) it('handles empty content array', async () => { @@ -338,10 +338,36 @@ describe('tool execution edge cases', () => { await syncTools(client as never, ctx, defaultOpts, new Map()) const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'empty_tool', arguments: {} }) - expect(result.content[0]).toEqual({ type: 'text', text: '(empty_tool returned no text content)' }) + expect(result.content[0]).toEqual({ type: 'text', text: '(empty_tool returned no content)' }) }) + it('uses fallback error message when isError with empty content', async () => { + const client = createMockClient( + [{ name: 'empty_err', inputSchema: { type: 'object' } }], + { content: [], isError: true }, + ) + + await syncTools(client as never, ctx, defaultOpts, new Map()) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'empty_err', arguments: {} }) + + expect(result.isError).toBe(true) + expect(result.content[0]).toEqual({ type: 'text', text: 'Error: MCP tool error' }) + }) + + it('surfaces structuredContent when content array is empty', async () => { + const client = createMockClient( + [{ name: 'structured', inputSchema: { type: 'object' } }], + ) + client.callTool.mockResolvedValue({ content: [], structuredContent: { key: 'value', count: 42 } }) + + await syncTools(client as never, ctx, defaultOpts, new Map()) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'structured', arguments: {} }) + + expect(result.isError).toBe(false) + expect(result.content[0]).toEqual({ type: 'text', text: '{"key":"value","count":42}' }) + }) + it('handles legacy toolResult with undefined value', async () => { const client = createMockClient( [{ name: 'legacy2', inputSchema: { type: 'object' } }], From 376d405ba48faf6a9f973d002d8456988d0addf8 Mon Sep 17 00:00:00 2001 From: Tianyi Cui <53024+tianyicui@users.noreply.github.com> Date: Wed, 8 Jul 2026 00:45:21 +0800 Subject: [PATCH 09/26] =?UTF-8?q?docs:=20rewrite=20the=20Code=20Mode=20RFC?= =?UTF-8?q?=20=E2=80=94=20registry-native=20mode=20over=20a=20worker-threa?= =?UTF-8?q?d=20code-runtime=20seam?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Supersedes the add-on-plugin + node:vm-stub draft in place (still proposed/): code mode becomes a first-class ToolRegistry presentation mode ('native' | 'code' | 'both'), execution goes behind a new ctx.codeRuntime capability seam whose shipped backend is one fresh Node worker thread per run (type-strip, empty env, resource limits, hard terminate), at bash-equivalent trust with no unsafe-flag ceremony. Renames the file to 2026-06-15-code-mode.md and regenerates the RFC index. --- docs/rfc/INDEX.md | 2 +- .../proposed/feature/2026-06-15-code-mode.md | 137 ++++++++++++++++++ .../feature/2026-06-15-optional-code-mode.md | 119 --------------- 3 files changed, 138 insertions(+), 120 deletions(-) create mode 100644 docs/rfc/proposed/feature/2026-06-15-code-mode.md delete mode 100644 docs/rfc/proposed/feature/2026-06-15-optional-code-mode.md diff --git a/docs/rfc/INDEX.md b/docs/rfc/INDEX.md index 2a496e469b..df5cf938c5 100644 --- a/docs/rfc/INDEX.md +++ b/docs/rfc/INDEX.md @@ -10,7 +10,7 @@ Generated by `pnpm run gen-rfc-index` from the RFC tree — never edit by hand; |---|---| | [Agent Client Protocol (ACP) support — drive the coding agent from external editors](proposed/feature/2026-06-14-acp-agent-client-protocol.md) | 2026-06-14 | | [Multiplex concurrent ACP sessions over one connection](proposed/feature/2026-06-14-acp-multi-session.md) | 2026-06-14 | -| [Optional Code Mode — model writes TypeScript against an SDK of all tools](proposed/feature/2026-06-15-optional-code-mode.md) | 2026-06-15 | +| [Code Mode — the model writes TypeScript against the tool registry](proposed/feature/2026-06-15-code-mode.md) | 2026-06-15 | | [Pre-tool input rewrite — a consistent design](proposed/feature/2026-06-30-pre-tool-input-rewrite.md) | 2026-06-30 | | [Claude Code and Codex subagent backends (out-of-process delegation to external coding agents)](proposed/feature/2026-07-07-claude-code-and-codex-subagent-backends.md) | 2026-07-07 | | [MCP client plugin — connect to external MCP servers and bridge their tools](proposed/feature/2026-07-07-mcp-client-plugin.md) | 2026-07-07 | diff --git a/docs/rfc/proposed/feature/2026-06-15-code-mode.md b/docs/rfc/proposed/feature/2026-06-15-code-mode.md new file mode 100644 index 0000000000..d868c41faf --- /dev/null +++ b/docs/rfc/proposed/feature/2026-06-15-code-mode.md @@ -0,0 +1,137 @@ +# RFC: Code Mode — the model writes TypeScript against the tool registry + +Status: proposed + +## Problem + +Today the agent loop advertises every registered tool to the model as a native JSON-schema function definition. `ToolRegistry` contributes its schemas to the system-prompt assembly, the assembly's `tools` land on the wire (and in the logged request header), the model invokes one `tool-call` block per step, and the loop dispatches each call through `ctx.tools.execute()` **sequentially** (parallel tool execution is an explicit open TODO in `dsh-tools` and [docs/architecture.md](../../../architecture.md)), with **every** intermediate `tool-result` re-entering the model's context on the next request. + +For multi-step tool work this is token-heavy and serial. The model cannot compose tools — loop over a result set, branch on an intermediate value, fan out, post-process — without a full model round-trip per call, and each round-trip drags the entire intermediate result back into context whether the model needs it or not. + +Cloudflare's [Code Mode](https://blog.cloudflare.com/code-mode/) proposes an alternative grounded in a simple observation: LLMs are better at writing code than at emitting tool calls, because they have seen millions of lines of real code and comparatively few contrived tool-calling traces. Instead of one tool call per step, the model writes a TypeScript program against a generated API over the tools, the program executes in a sandboxed runtime, and the model curates what comes back — only what it prints or returns — instead of every intermediate result. + +An earlier draft of this RFC designed Code Mode as an add-on consumer plugin with zero core changes, deferring the execution substrate to a follow-up. Both constraints are dropped here, deliberately. First, the harness is pre-release and optimizes for the correct foundation over blast radius: tool presentation is the registry's own concern, and bolting a second presentation onto it from outside means transforming the registry's contribution after the fact — a waterfall listener whose correctness depends on listener ordering, which fights the [reconstructable-requests](../../implemented/architecture/2026-07-05-reconstructable-requests.md) design instead of riding it (that refactor removed request mutation from `agent/request`, the seam the old draft relied on). Second, the substrate question is answerable now: a Node `worker_threads` runtime gives real containment — separate isolate, empty environment, heap caps, and a `terminate()` that reliably stops a hot synchronous loop — where the old draft's `node:vm` stub had none of those, and it fits the harness's existing trust model (§Trust posture) without a hardening follow-up. + +## Proposal + +Three decisions, each elaborated in its own section below: + +1. **Code Mode is a first-class presentation mode of `ToolRegistry`** (`dsh-tools`), selected by a validated `mode` config: `'native'` (today's behavior, the default), `'code'` (the wire carries exactly one tool, `run_code`, plus a generated SDK `.d.ts` in the system prompt), or `'both'` (native schemas *and* `run_code` + SDK). The registry's existing tool-schema provider contributes whatever the mode dictates, so the wire tool list is shaped at its source — no interception, no listener-ordering caveats — and the logged request header records it for free. +2. **Code execution is a capability seam** — a new group `packages/code-runtime/` with the interface package `@deepseek-ai/dsh-code-runtime` owning `ctx.codeRuntime` ([capability seams](../../implemented/architecture/2026-06-13-capability-seams.md); consumer = `dsh-tools`, with core-consumes-a-seam precedent in `agent-loop` → `dsh-llm`). The runtime knows nothing about tools: it is handed a program and named async bindings, runs the program, and reports `{ value, logs, error? }`. Language and substrate are backend properties, so a future Python or container backend is a new implementation package, not a redesign. +3. **The shipped implementation is `@deepseek-ai/dsh-code-runtime-worker`**: one fresh Node worker thread per run, executing the model's TypeScript after type-strip, with bindings bridged over the message port, an empty environment, configurable heap/output/time caps, and hard termination. Its trust posture is bash-equivalent by design — no unsafe-acknowledgement flags — because the harness already ships `dsh-bash-local`, which executes arbitrary model-written shell commands with strictly *more* ambient authority. + +### The registry owns the mode + +`ToolRegistry` gains a schemastery-validated config (`static Config`), its first: `mode: 'native' | 'code' | 'both'`, default `'native'`. A deployment flips it from `cordis.yml` (`tools: { mode: code }`) — no code edit, per the no-hardcoded-tunables convention. + +**Wire tool list = the provider's contribution.** The registry already feeds the assembly through `ctx.systemPrompt.tools(() => this.schemas())`; the provider becomes mode-aware: `'native'` contributes all schemas (unchanged), `'code'` contributes only `run_code`'s schema, `'both'` contributes all schemas plus `run_code`. Because [`PromptAssembly.tools` is the single source the loop's request header snapshots](../../../../packages/core/system-prompt/src/index.ts), the collapse is automatically logged and reconstructable — model-visible ⟺ logged holds with zero new mechanism, and there is no seam left where another party could accidentally re-add tools (the `system-prompt/assemble` waterfall can still deliberately transform the assembly, as ever). + +**Interaction with `toolOrder`, stated up front:** a configured `systemPrompt.toolOrder` naming native tools rejects every assembly under `mode: 'code'` (those names are no longer contributed), by the existing fail-loud rule for unlisted names. This is correct behavior, not a bug: a deployment switching modes updates its order config or drops it. + +**The SDK prompt section.** Under `'code'` and `'both'` the registry registers one lazy prompt section (`tools:sdk`, in the 100–199 tool-guidance order band) whose thunk regenerates, at each assembly, a TypeScript declaration of every registered tool except `run_code` itself, plus fixed usage instructions. The thunk reads the live store and emits tools in lexicographic name order, so its output is deterministic and stable across steps — an unchanged tool set produces byte-identical text (prefix-cache-friendly; a mid-session registration surfaces as one logged header delta, exactly like a native-mode tool change). + +**Codegen.** A pure `jsonSchemaToTs(schema)` module inside `dsh-tools` (sibling of `json-schema.ts` — `schemas()` and the SDK are two projections of the same store) maps the JSON-Schema subset the `defineTool` DSL emits (object/string/number/boolean/array, `properties`, `required`, string `enum` → literal union, nested objects, array `items`, `description` → JSDoc) to a TS type literal. It is **total**: any construct outside that subset (`$ref`, `oneOf`/`anyOf`, `integer`, future MCP shapes, …) degrades to `unknown` without throwing. Because `ToolSchema.name` is an arbitrary string, the SDK is declared as one object constant — `declare const tools: { "some-mcp-tool"(args: …): Promise; bash(args: …): Promise; … }` — quoted keys make every name reachable with no sanitization or alias-collision logic. Typing is advisory (the runtime executes type-stripped JS); the instructions say so. + +### The run_code tool and the dispatch bridge + +Under `'code'` and `'both'` the registry registers `run_code` in itself as an ordinary tool — one required parameter `{ code: string }` — so the unchanged loop dispatches it through the normal pipeline and `tools/pre-execute` / `tools/post-execute` gate it like any other call (a permission plugin can inspect the program text before it runs). Its `execute(args, exec)`: + +1. **Builds the bindings**: for every registered tool except `run_code`, an async function that (a) checks `exec.signal?.aborted` before and after (throwing stops the program — necessary because `ctx.tools.execute()` converts errors to `isError` data), (b) awaits its turn on the **per-run serialization queue** (below), (c) calls `this.execute({ callId, name, arguments, agent: exec.agent, signal: exec.signal })` with a deterministic sub-id `` CallId(`${exec.callId}:code:${n}`) ``, (d) appends a `tool/code-dispatch` session event, and (e) maps the result: success → the text-block contents joined as a `string` (non-text blocks become placeholders, an MVP limitation), `isError` → **the binding rejects** with an `Error` carrying the result text. Rejection is the deliberate model-facing contract — real code signals failure by throwing, `try/catch` and `Promise.all` short-circuiting behave as every model has seen them behave — where the old draft's `{ output, isError }` envelope made error handling a bespoke convention. +2. **Runs the program**: `ctx.codeRuntime.run({ program: args.code, bindings: [{ global: 'tools', functions }], signal: exec.signal })`. +3. **Surfaces the outcome**: a successful run returns one text block — the captured console/stdout output followed by the rendered return value (if any) — plus a `meta` payload (capped logs, dispatch count) for presentation. A run with `result.error` throws a `CodeRunFailedError extends HarnessError` (`code: 'CODE_RUN_FAILED'`, message = the error kind and text plus captured logs so the model can self-correct); the registry's existing catch turns it into a structured `isError` result. + +**Concurrency: serialized, enforced by the binding.** The bindings are async, so a model writing `Promise.all([tools.a(…), tools.b(…)])` starts both immediately — concurrent dispatch would be the *default*, while the tool contract still carries no concurrency-safety metadata (the open parallel-execution TODO). Each `run_code` invocation therefore owns a dispatch queue and every binding call chains onto it, so even `Promise.all` executes the underlying `ctx.tools.execute()` calls one at a time in submission order; when the run settles, queued-but-unstarted dispatches are abandoned. Lifting this per-tool once tools can declare themselves concurrency-safe is deferred work, same as before. + +**Presentation.** `run_code`'s render intent is decided here per the [render-intent RFC](../../implemented/architecture/2026-07-02-tool-render-intent-union.md): `presentCall` → a `generic` card, `kind: 'execute'`, title `Run code`, `rawInput` = the program text; `presentResult` → a `generic` card whose content is the captured output (from `meta`). Not a `terminal` card: that card's semantics are "a shell command in a working directory", which a program is not. + +### Observability: `tool/code-dispatch` + +Each sub-dispatch appends one session event, declared by `dsh-tools` via `SessionEventMap` declaration merging (the map is merge-extensible for exactly this): `tool/code-dispatch` with `{ parentCallId, subCallId, name, arguments, isError, resultSummary }`. It is log-only — `deriveEventMessage()` ignores unknown event types by design, so sub-calls never re-enter model context — but persistence and UIs get every call. A `run_code` execution arriving without `exec.agent` (the loop always supplies it; direct programmatic calls may not) still runs and simply skips event logging, exactly as the `ToolExecution` contract allows. + +### The code-runtime seam + +`packages/code-runtime/code-runtime/` — `@deepseek-ai/dsh-code-runtime`, depending only on `cordis`. An abstract `CodeRuntime extends Service` (`super(ctx, 'codeRuntime')`) plus the vocabulary: + +- `CodeRunRequest = { program: string; bindings: CodeBindingNamespace[]; signal?: AbortSignal }` +- `CodeBindingNamespace = { global: string; functions: Record Promise> }` — the runtime exposes each namespace as a global object of async functions inside the program; binding arguments and resolutions must be structured-cloneable (a runtime may cross a serialization boundary; ours does). +- `CodeRunResult = { value?: unknown; logs: CodeLogEntry[]; error?: CodeRunFailure }` — an error is a field on a resolved result, never a rejection of `run()`. +- `CodeLogEntry = { source: 'console' | 'stdout' | 'stderr'; level?: 'log' | 'info' | 'warn' | 'error' | 'debug'; text: string }` +- `CodeRunFailure = { kind: 'exception' | 'timeout' | 'abort' | 'worker-exit'; message: string }` — orthogonal outcomes reported independently per [defensive patterns](../../../defensive-patterns.md); a timed-out run is not an exception, an abort is not a timeout. +- Two readonly backend descriptors, informational not gating: `language` (what the program must be written in — `'typescript'` for the shipped backend; a Python backend would say so, and pair with its own SDK generator on the presentation side) and `isolation` (`'worker-thread'` for the shipped backend; `'process'`, `'container'`, … for future ones). `dsh-tools` requires `language === 'typescript'` in the MVP — its codegen emits TS — and fails the assembly loudly otherwise, the same misconfiguration idiom as `toolOrder` violations (as when `mode` is non-native with no `ctx.codeRuntime` loaded at all). + +Per explicit-over-implicit at seams, the request spells out everything the runtime acts on; defaulting (timeouts, caps) is the implementation's validated config, never a hidden `??` inside `run()`. The seam split is justified by real planned divergence on both axes — substrate (worker now; container/microVM later) and language (the Python/AssemblyScript direction sketched in the earlier draft survives as future work) — not by speculation: `dsh-tools` consumes the interface today and tests against a trivial in-repo fake, exactly the interface/implementation/consumer shape of the bash template. + +### The worker-thread runtime + +`packages/code-runtime/code-runtime-worker/` — `@deepseek-ai/dsh-code-runtime-worker`. Per `run()`: + +1. **Type-strip host-side** with Node's built-in `stripTypeScriptTypes` (`node:module`; the repo floor is node ≥ 24, and the API is position-preserving, so runtime error line numbers match the model's source). Strip-only mode rejects non-erasable syntax (`enum`, namespaces) — that rejection returns as `error.kind: 'exception'` with Node's message, the SDK instructions say "erasable TypeScript only", and the model self-corrects like any other program error. A syntax-level failure never spawns a worker. +2. **Spawn one fresh `Worker` per run** from the package's own bootstrap module: `env: {}` (truly empty — stronger than the scrubbed-env rule for spawned commands), `resourceLimits` from config, `stdout`/`stderr` captured into `logs` rather than inherited. No pooling and no cross-run state: a program's world dies with its worker, which keeps runs reconstructable from the log alone and makes state bleed unrepresentable. +3. **Execute** in the bootstrap: the stripped program becomes the body of an `AsyncFunction` whose parameters are the binding globals and a capturing `console` shim, so top-level `await` and `return` work and the program's completion value is the run's `value` (structured-cloneable values cross as-is; anything else is replaced by its `util.inspect` rendering, documented). +4. **Bridge bindings over the message port**: each binding function in the worker posts `{ id, global, name, args }` and awaits the reply; the host validates the name against the request's bindings, invokes, and replies `{ id, ok, value }` or `{ id, ok: false, message }` (a host-side binding rejection becomes a program-side rejection). Unknown names, duplicate ids, and post-settlement messages are rejected or ignored — the port protocol assumes a hostile peer, because the peer runs model code. +5. **Enforce caps**: the compute-time budget (`timeoutMs`) ticks only while no binding call is pending — it bounds runaway worker compute, not tool latency — and expiry, `signal` abort, and run completion all funnel into `worker.terminate()`, which ends hot synchronous loops too (measured; this was `node:vm`'s unfixable gap). Heap overflow surfaces as the worker's OOM exit → `error.kind: 'worker-exit'`. Log and value sizes are capped by config (`maxLogBytes`, `maxValueBytes`), truncation marked in-band. All caps are validated config fields with defaults (`timeoutMs: 60_000`, `maxLogBytes: 65_536`, `maxValueBytes: 32_768`, `maxOldGenerationSizeMb: 512`), changeable from `cordis.yml`. +6. **Dispose to quiescence**: the service's own disposal terminates in-flight workers and *awaits* their exits before resolving, per [defensive patterns](../../../defensive-patterns.md). + +### Trust posture + +The worker runtime is **containment, not a security boundary**, and the RFC says so without ceremony. Model code in the worker can reach Node globals — `fetch`, `process` (with an empty env) — so a deliberately adversarial program has ambient authority comparable to what the harness's own `bash` tool already grants every model turn: `dsh-bash-local` runs arbitrary model-written commands with the host filesystem, network, and a scrubbed-but-populated environment. Code Mode is gated where bash is gated — `tools/pre-execute`, where permission/sandbox plugins veto or approve the program before it runs — and adds containment bash does not have: empty env, heap caps, hard termination, a separate isolate. The earlier draft's two-flag unsafe ceremony (`{ unsafe: true }` constructor + `allowUnsafeRuntime`) existed for a `node:vm` stub with *no* containment and is dropped with it; demanding scarier flags for the better-contained executor than for bash would be posture theater. A deployment that needs a hard boundary (untrusted multi-tenant input) needs it for bash too; that is a future `isolation: 'container'` backend, and the `isolation` descriptor exists so such a deployment can tell backends apart. + +### What the model sees + +The `tools:sdk` section carries the `.d.ts` plus fixed instructions: the program is the body of an async TypeScript function (erasable syntax only — no `enum`/namespaces; type annotations are advisory); call tools as `await tools.name(args)` (quoted access for exotic names); a failed tool call **rejects** with an `Error` carrying the tool's error text — catch it to handle and continue; calls run **sequentially** even under `Promise.all`; emit results via `return` and/or `console.log`, and only that curated output returns to the context — intermediate tool results never do. That last line is the payoff the whole design serves: output-side context cost becomes the model's own editorial decision. On the input side the `.d.ts` is not free — for a large tool surface it can rival the native JSON schemas it replaces (and `'both'` pays for the two side by side) — but it is prefix-stable, so provider prefix caching amortizes it; the win is workload-dependent and the RFC claims no more. + +## Plan + +Four stacked PRs, each gates-green (`typecheck`, `lint`, `test:coverage`, `test:snapshot`, `doc-sync`, `verify-module-graph`, `build`, `hygiene`) with docs updated in the same change: + +1. **This RFC revision** (docs-only): the file rewritten as above (renamed `2026-06-15-code-mode.md`, same first-proposed date), regenerated RFC index. +2. **`dsh-code-runtime`** (interface package): the group `packages/code-runtime/`, abstract `CodeRuntime`, vocabulary types, ctx-key declaration, group README + package README + catalog updates. Unit tier: HMR safety (dispose removes `ctx.codeRuntime`), contract docs. This package has no behavior to snapshot or e2e; its coverage story is unit-only by design. +3. **`dsh-code-runtime-worker`**: the implementation above. Unit tier (real workers, no mocks — they are cheap and local): output/value capture, log source attribution, error kinds (exception incl. non-erasable syntax, timeout vs abort vs worker-exit under OOM), compute-time-only timeout (a slow binding does not expire the run), binding bridge hostility cases (unknown name, duplicate id, post-settlement message), structured-clone fallback, cap truncation, `env` emptiness verified from inside the program, dispose-awaits-exit. A real-load-path test runs the built package (`lib/`) so the worker entry resolves both unbuilt (tsx) and built — the published-bin guard from [docs/testing.md](../../../testing.md). +4. **Native code mode in `dsh-tools`** + the end-to-end surface: mode config, provider switch, `tools:sdk` section, `jsonSchemaToTs`, `run_code` + dispatch bridge + serialization queue, `tool/code-dispatch` event; [docs/architecture.md](../../../architecture.md) (tool-pipeline section + `ctx.codeRuntime` service row) and the [adding-a-tool cookbook](../../../cookbook/adding-a-tool.md) cross-reference updated in the same change; an `examples/` leaf + `demo:code` script wiring the worker runtime with `mode: 'code'`; move this RFC to `implemented/`. Coverage named per tier now, per the plan-time rule: **unit** — codegen table (DSL subset, quoted names, `unknown` degradation, determinism), provider contribution per mode, `toolOrder × mode` rejection, missing-runtime/wrong-language loud failures, serialization non-overlap (a probe tool records enter/exit under `Promise.all`), abort stops the queue, binding rejection on `isError`, `CodeRunFailedError` → structured `isError`, event payloads, `deriveMessages()` ignores the event, HMR safety (mode flip via config reload removes tool + section); **e2e (with-key, self-skips)** — a real model, `mode: 'code'`, a task requiring two tool calls and curation, asserting the wire tool list was exactly `run_code` and the transcript's dispatch events; **snapshot (keyless replay)** — goldens for a `run_code` turn in `'code'` and `'both'`, pinning the SDK section text, the collapsed header tools, dispatch events, and the result card. + +The four PRs land in order (each on the previous); per stacked-review practice, review fixes land on the PR that introduced them and merge down. + +## Alternatives considered + +**An add-on consumer plugin, zero core changes (the previous draft of this RFC).** Rejected on both halves. The wire-collapse half aged out from under it: it targeted the `agent/request` waterfall, which [reconstructable requests](../../implemented/architecture/2026-07-05-reconstructable-requests.md) has since re-typed to call-config-only, and the surviving alternative — transforming the assembly a waterfall listener receives — is strictly worse than contributing the right list in the first place (transformation must undo `toolOrder` canonicalization it cannot see the config for, and its correctness depends on where it sits in a listener chain). The deeper reason is ownership: which tools the model is offered, in which representation, is the registry's single concern — `schemas()` for function calling and the SDK for Code Mode are two projections of one store, and splitting the second projection into a satellite package would preserve a boundary the domain does not have. + +**`node:vm` as the reference runtime, hardening deferred (also the previous draft).** Rejected: `node:vm` is not isolation (prototype-chain escapes reach the host realm), cannot interrupt a hot loop, and forced the draft into a two-flag unsafe ceremony plus a mandatory follow-up RFC. The worker thread delivers the missing properties now — separate isolate, empty env, `resourceLimits`, reliable `terminate()` (all verified by probe before this revision) — at bash-equivalent trust, so the reference implementation and the production one are the same package and the ceremony dissolves. + +**Result elision / summarization over native tool-calling.** Addresses only the context-bloat half of the problem: trimming old `tool-result`s (now cheap to add as a logged surface replace, per the reconstructable-requests consequences) still pays one model round-trip per call and cannot express loops, branches, or joins. Complementary, not competing; it can layer under Code Mode for residual native calls. + +**Parallel native dispatch in the loop.** The other answer to round-trip cost; still valid future work (the open TODO), still blocked on concurrency-safety metadata, and still no composition — it parallelizes calls the model already decided on in one step. Code Mode's serialized-queue decision keeps the two compatible: when the metadata lands, both native parallel dispatch and per-tool binding parallelism unlock together. + +**Always-exclusive (Cloudflare-faithful, no mode).** Rejected for this SDK's primary consumer: a coding agent's bread-and-butter single calls (`bash`, `read`, `edit`) are already ideal as native calls, and forcing every edit through a program taxes the common case. The mode config keeps the faithful form (`'code'`) one line away without imposing it. + +**Per-tool visibility tiers (this tool native, that tool code-only).** Deferred again, knowingly: it needs per-tool metadata and a presentation split that `'native' | 'code' | 'both'` does not, and every learning it depends on (how models actually split usage under `'both'`) arrives only after this ships. + +**Sanitized identifier aliases in the SDK** (`my-tool` → `my_tool`, Cloudflare's approach). Rejected: quoted keys on a `declare const` make every name reachable with zero alias-collision logic; models handle `tools["my-tool"](…)` fine. + +**A REPL-style persistent kernel** (state survives across `run_code` calls). Rejected for the MVP: cross-call state would be invisible to the session log, breaking the reconstructability guarantee that every request is a pure function of the log; fresh-per-run keeps it. A kernel-style backend remains expressible behind the seam later, with its own logging story. + +## Acceptance criteria + +- `mode: 'native'` (and unset) is byte-for-byte today's behavior: same assemblies, same headers, same snapshots. +- Under `mode: 'code'`, the assembled tool list (and thus the logged `request/header`) is exactly `[run_code]`; under `'both'`, every native schema plus `run_code`; the `tools:sdk` section is present in both, absent under `'native'`, and its text is deterministic for a fixed tool set (byte-identical across consecutive assemblies). +- The generated `.d.ts` covers every registered tool except `run_code`, non-identifier names via quoted keys, unsupported schema constructs as `unknown`, without codegen ever throwing. +- A program calling two tools returns only its curated output; each sub-call appears as a `tool/code-dispatch` event ordered by log `seq`, flows through `tools/pre-execute`/`post-execute` (a deny reaches the program as a binding rejection), and never enters derived messages. +- `Promise.all` over three bindings produces non-overlapping `ctx.tools.execute()` intervals (probe-tool assertion); aborting mid-program stops the worker and dispatches nothing further. +- Worker runtime: a hot `for(;;){}` run ends at the compute-time budget with `error.kind: 'timeout'`; a pending binding call does not consume that budget; OOM under `resourceLimits` yields `kind: 'worker-exit'` with the host process healthy; `process.env` inside a program is empty; non-erasable syntax yields `kind: 'exception'` without a worker spawn; disposal awaits worker exit. +- Misconfiguration is loud before any model request: non-native mode with no `ctx.codeRuntime`, a runtime whose `language ≠ 'typescript'`, and `toolOrder` naming a non-contributed tool all reject the assembly with actionable messages. +- The demo runs against the real API via `demo:code`; the snapshot goldens replay keylessly; all repo gates pass on every PR of the stack. + +## Risks + +**The worker is not a hard security boundary.** Deliberate and documented (§Trust posture): posture equals the existing bash tool, containment exceeds it, gating uses the same seams. Deployments needing more need a future `isolation: 'container'` backend — tracked as the seam's designed extension, not a TODO on this design. + +**`stripTypeScriptTypes` is marked experimental.** It is also what Node itself runs `.ts` files with on this repo's node floor. Mitigations: the runtime's unit suite pins the behaviors relied on (position preservation, erasable-only rejection message shape loosely), the call sits behind one private function, and `amaro`/`sucrase` are drop-in replacements if the API shifts. The erasable-only subset is a model-facing contract line, and the error path is a working feedback loop, not a dead end. + +**Prompt cost of the SDK, especially under `'both'`.** The `.d.ts` can rival the native schemas it complements; `'both'` carries two representations. Prefix stability + provider caching amortize per-session cost; the mode is per-deployment; the RFC makes no unconditional-savings claim. Measured guidance (when to prefer which mode) is explicitly post-ship learning. + +**Registry scope growth.** `dsh-tools` absorbs codegen, a tool, a bridge, and an event. Contained by module boundaries inside the package (`ts-types.ts`, `code-mode.ts` beside `schema.ts`/`json-schema.ts`/`presentation.ts`) and by the seam: everything substrate-shaped lives behind `ctx.codeRuntime`. + +**Structured-clone limits at the binding boundary.** Tool bindings pass JSON-shaped arguments and return strings in the MVP, comfortably cloneable; the seam contract states the constraint so a future binding producer cannot discover it in production. Non-text sub-result content is reduced to placeholders — a known MVP limitation, recorded in the SDK instructions. + +**Serialized-only sub-dispatch.** `Promise.all` gains no wall-clock parallelism yet, only fewer round-trips; models may over-expect. The instructions state it; lifting it is tied to the same concurrency-safety metadata the native parallel-dispatch TODO needs. + +**Timer-pause bookkeeping.** Compute-time-only budgeting adds a small state machine (pause on RPC out, resume on reply) whose bugs would misattribute time. It is unit-tested from both sides (slow binding ≠ timeout; hot loop = timeout) and is still simpler than the alternative — one budget that spuriously kills programs for waiting on a slow tool. diff --git a/docs/rfc/proposed/feature/2026-06-15-optional-code-mode.md b/docs/rfc/proposed/feature/2026-06-15-optional-code-mode.md deleted file mode 100644 index e221618ce1..0000000000 --- a/docs/rfc/proposed/feature/2026-06-15-optional-code-mode.md +++ /dev/null @@ -1,119 +0,0 @@ -# RFC: Optional Code Mode — model writes TypeScript against an SDK of all tools - -Status: proposed - -> Premise partially stale: this proposal predates [request reconstructability](../../implemented/architecture/2026-07-05-reconstructable-requests.md) — `agent/request` now shapes call config only (no request/content mutation), so the interception points named below need re-mapping onto the log channels and `system-prompt/assemble` before implementation. - -## Problem - -Today the agent loop advertises every registered tool to the model as a native JSON-schema function definition. `ToolRegistry` feeds its schemas into `ctx.systemPrompt`, the loop puts them on `GenerateOptions.tools`, and the adapter serializes them to the provider's function-calling wire format. The model then invokes one `tool-call` block per step, the loop dispatches each call through `ctx.tools.execute()` **sequentially** (parallel tool execution is an explicit open TODO in `dsh-tools` and [docs/architecture.md](../../../architecture.md)), and **every** intermediate `tool-result` re-enters the model's context on the next request. - -For multi-step tool work this is token-heavy and serial. The model cannot compose tools — loop over a result set, branch on an intermediate value, fan out, post-process — without a full model round-trip per call, and each of those round-trips drags the entire intermediate result back into context whether the model needs it or not. - -Cloudflare's [Code Mode](https://blog.cloudflare.com/code-mode/) (shipped as the `@cloudflare/codemode` npm package) proposes an alternative grounded in a simple observation: LLMs are better at writing code than at emitting tool calls, because they have seen millions of lines of real code and comparatively few contrived tool-calling traces. Instead of one tool call per step, the model writes a TypeScript program against a generated SDK that wraps all the tools, and that program is executed. The model curates what comes back — only what it `console.log`s and/or returns — instead of every intermediate result. The SDK functions are async, so the model can *express* fan-out (`Promise.all`) naturally in code; this RFC initially **serializes** those dispatches (§ Concurrency) until the tool contract grows concurrency-safety metadata, so the early win is composition and fewer round-trips, not parallelism. - -This RFC proposes an **optional** Code Mode for the DeepSeek Harness, covering **all** tools uniformly — built-in and future MCP — with no per-tool work, implemented Cordis-style with **zero core-package changes**. It fully specifies the code-execution seam and the SDK-generation pipeline, but ships only a minimal `node:vm` reference stub for execution; the hardened, sandboxed execution substrate is **deferred to a follow-up RFC** (see Risks). This RFC does not change the agent loop, and it leaves native tool-calling exactly as it is — Code Mode is a plugin you load, not a replacement. - -## Proposal - -The design follows the codebase's capability-seam pattern ([capability seams](../../implemented/architecture/2026-06-13-capability-seams.md), the `bash` template) as a three-package split, plus one consumer plugin. Nothing in `dsh-session`, `dsh-agent`, `dsh-agent-loop`, `dsh-llm`, `dsh-tools`, or `dsh-system-prompt` changes. - -**Prior art.** `@cloudflare/codemode` validates this shape directly and several of its decisions are adopted below. Its `Executor` interface is deliberately tiny — `execute(code, fns) → { result, error?, logs? }` — with a production `DynamicWorkerExecutor` (isolated Workers) and a six-line `NodeVMExecutor` example as two implementations behind it: exactly the interface/implementation split [the capability-seam pattern](../../implemented/architecture/2026-06-13-capability-seams.md) prescribes. It generates TypeScript type definitions from tools for the model's context and runs the generated JavaScript in a sandbox, capturing console output alongside the return value. It normalizes model output into an async arrow function via AST parsing (acorn) and sanitizes tool names into valid JS identifiers (`my-tool` → `my_tool`, `delete` → `delete_`). It blocks outbound network by default. The transferable lessons — minimal executor contract, host-side type derivation, capture-output-and-return-value, name sanitization, AST-normalize the code, isolate by default — are folded into the design below. What does **not** transfer is the substrate: Cloudflare's isolation is Workers-specific; our equivalent hardened substrate is the deferred follow-up. - -**Prompt-budget tradeoff (Code Mode is not unconditionally cheaper).** Deriving the SDK types host-side costs no extra *discovery* round-trip, but the generated `.d.ts` is injected into the system prompt (§3a), so the type definitions themselves **do** consume context — and for an all-tools SDK that cost scales with every registered tool and can be comparable to, or larger than, the native JSON schemas it replaces. Code Mode's saving is on the **output/result** side (the model curates what comes back; intermediate results never re-enter context) and on **round-trips** (compose many calls in one program), not on the input-side tool description. The net win is workload-dependent: it pays off for multi-call, large-intermediate-result workflows and can cost *more* for a single call against a large tool surface. The `.d.ts` section is a prefix-stable prompt prefix, so prompt caching amortizes its per-turn cost across a session; the RFC notes that caching is what keeps the injected SDK affordable, and that a deployment with a very large tool surface should weigh the SDK size against native schemas rather than assume Code Mode is strictly cheaper. - -**1. Interface package `packages/code-runtime/`** — a new package `@deepseek-ai/dsh-code-runtime` owning `ctx.codeRuntime`, depending only on `cordis`. It defines an abstract `CodeRuntime extends Service` plus the execution vocabulary. The runtime knows **nothing** about `ctx.tools`: it is handed a set of named async functions (the resolved SDK bindings), runs the program, and captures output. The result shape mirrors Cloudflare's proven-minimal contract so an error is a *field on a resolved result*, not a throw the runtime is expected to make: - -- `CodeRunRequest = { code: string; sdk: SdkBinding[]; signal?: AbortSignal }` -- `CodeRunResult = { result: unknown; logs: string[]; error?: string }` -- a readonly `safe: boolean` on the `CodeRuntime` service — `false` for an unsandboxed stub, `true` only for a real isolating substrate; consumers gate on it (§2). -- `SdkBinding = { namespace: string; fns: Record Promise> }` - -Per the "explicit > implicit at seams" convention, the request spells out every field the runtime acts on; defaulting (e.g. an output cap, a timeout derived from `signal`) is the implementation's explicit job, not a hidden `?? default` inside `run()`. The split into interface + implementation is justified under [the capability-seam pattern](../../implemented/architecture/2026-06-13-capability-seams.md) because there is **genuinely more than one planned implementation** — the node:vm stub *and* the hardened substrate (a real isolate, or the generated program run as a sandboxed process through the existing `ctx.bash` seam) that is scheduled follow-up work, not speculative optionality. The capability-seam pattern warns against splitting preemptively when only one implementation is conceivable; here a second is not just conceivable but required before any untrusted use, so the seam earns its keep. - -**Backends can differ by language/runtime, not only by trust level.** The two implementations above (unsafe stub vs. hardened substrate) differ along the *trust* axis while staying TypeScript/JS, but nothing in the `CodeRuntime` contract — a program string plus a set of named async SDK bindings in, and a `{ result, logs, error? }` out — is bound to one source language. The same seam can host backends that differ along the *language* axis, executing a program written in something other than TypeScript. Two illustrative directions: - -- **An AssemblyScript backend.** AssemblyScript is a strict TypeScript subset that compiles to WebAssembly, so a program stays familiar to a TS-fluent model while the WASM boundary supplies exactly the sandboxing the hardened substrate is meant to provide — memory isolation and no ambient host authority come from the runtime rather than from after-the-fact hardening of `node:vm`. This is an appealing route to a `safe = true` backend. -- **A Python backend.** Python is arguably the model's most native language — it has seen far more real Python than any tool-calling trace — which is the same "LLMs write better code than tool calls" argument that motivates Code Mode, taken one step further. A Python backend is itself a sub-seam over different Python *runtimes*: **CPython** (in-process or a sandboxed subprocess via `ctx.bash`) for maximum fidelity and ecosystem access, or a more controllable / embeddable interpreter — Pyodide (CPython on WASM), RustPython, or a restricted embedded interpreter — when isolation, deterministic resource limits, or a clean capability boundary matter more than running arbitrary native extensions. - -These are illustrations of the seam's reach, **not commitments** — the MVP ships only the TypeScript path. The honest caveat is that the *execution* contract is language-agnostic but the *presentation* is not: the SDK-generation pipeline below (§3a and the `jsonSchemaToTs` codegen, which emits a TypeScript `.d.ts`) is TypeScript-specific, so a non-TS backend pairs the shared `CodeRuntime` contract with its own language-appropriate SDK generator and system-prompt section (a `.pyi` stub and Python usage instructions for the Python backend, AssemblyScript-flavored types for that one). The runtime seam is reused as-is; only the codegen/prompt half is per-language. - -**2. Implementation package `packages/code-runtime-vm/`** — a new package `@deepseek-ai/dsh-code-runtime-vm`, the `node:vm` reference stub. It type-erases the model's TypeScript via the compiler's `transpileModule` (or sucrase) — the types exist only to guide the model; the runtime is plain JS — then wraps the body in an async IIFE for top-level `await` (Cloudflare's `NodeVMExecutor` does literally `new AsyncFunction("codemode", "return await (${code})()")`), runs it in a `vm.Context` whose globals are a capturing `console` and the SDK namespace objects, awaits the IIFE, and captures the return value, the buffered logs, and any thrown error (as `error: string`). It applies an **output cap** (truncate captured logs) and a **timeout tied to `request.signal`**. These caps limit blast radius; **they are not a security boundary**. node:vm is **not** isolation: withholding `require`/`process` does not contain anything (code escapes via `constructor`/prototype reflection), and per [AGENTS.md](../../../../AGENTS.md) the harness must never hand model output the ambient environment. - -**The unsafe-runtime guard is enforceable, not a README warning.** Because a README caveat is not a control — and AGENTS.md's "never hand model output ambient authority" is a hard rule, not advice — the design makes the danger refuse to run by construction. Two layers: - -- **The runtime declares its trust level.** `CodeRuntime` carries a readonly `safe: boolean` (a `node:vm`-class stub returns `safe = false`; a real isolate/sandboxed-process substrate returns `safe = true`). The `code-runtime-vm` constructor *additionally* requires an explicit opt-in — `new VmCodeRuntime({ unsafe: true })` — and **throws** if that flag is absent, so merely depending on the package and wiring it cannot silently produce a live unsafe runtime; the operator must type the word `unsafe`. -- **The consumer refuses to expose `run_code` over an unsafe runtime by default.** When `code-mode` initializes, if `ctx.codeRuntime.safe === false` it does **not** register `run_code` unless the plugin itself is configured with an explicit acknowledgement (e.g. `code-mode` config `allowUnsafeRuntime: true`). Absent that, it logs a typed error and registers nothing — so a real model never reaches an unsandboxed runtime by a single config slip. The refusal path is tested: with the acknowledgement unset and an unsafe runtime, `run_code` is absent (and the wire tool list is unchanged from native); with both opt-ins set, it registers and runs. This keeps the unsafe reference backend usable for tests and trusted local demos while making production misuse take two deliberate, greppable flags rather than one mistake. - -`code-runtime-vm` is therefore documented as **reference / test-only / unsafe-for-untrusted-input**, acceptable in the MVP only because the code runs at harness trust *and* both opt-in flags must be set. Signal handling is best-effort: it aborts in-flight sub-dispatches but cannot reliably interrupt a hot synchronous loop (`while(true){}`) in node:vm — another reason the hardened substrate is deferred, not optional-forever. - -**3. Consumer plugin `packages/code-mode/`** — a new package `@deepseek-ai/dsh-code-mode`, the plugin that wires everything together. It declares `inject = ['tools', 'systemPrompt', 'codeRuntime']` — Cordis throws on access to a service that is not injected, and keeps the plugin inactive until all three exist (the same pattern as `tool-bash`'s `inject = ['tools', 'bash']`), which also gives correct load-ordering relative to `code-runtime`/`code-runtime-vm`. The plugin contributes four things, all through existing seams: - -**3a. Tool presentation — a lazy system-prompt section (the injection seam already exists).** `dsh-system-prompt` already provides the Cordis-idiomatic way for any plugin to inject prompt snippets: `ctx.systemPrompt.section({ name, order, text })`, fiber-scoped and auto-disposed via `ctx.effect()`, where `text` may be a lazy `() => string` re-evaluated at each assembly. No new mechanism is needed or invented. Code Mode registers a lazy section (high `order` so it lands last) whose thunk reads `ctx.tools.schemas()` at assembly time and regenerates the SDK `.d.ts` plus usage instructions from the currently-registered tool set. Because the thunk reads the live registry, coverage of every tool — built-in, MCP, future — is automatic. - -**3b. Wire tool-list enforcement — an `agent/request` listener (the authoritative seam).** The goal "exactly one tool reaches the wire" must be enforced where the wire request is finalized. The loop calls `ctx.systemPrompt.assemble()` first, *then* builds `GenerateOptions` (seeding `tools` from `assembly.tools`), *then* runs the `agent/request` waterfall, *then* calls `ctx.llm.stream()`. A `system-prompt/assemble` listener can only influence the *seed*; `agent/request` is the last seam before the model call, so it is authoritative. The plugin registers an `agent/request` listener that does `const final = await next(); return { ...final, tools: [runCodeSchema] }` — overriding the value *returned by* `next()`, not the inbound argument, so it dominates the cooperative request listeners it wraps. It registers with `prepend: true` to sit at the outer edge of the waterfall chain. One honest caveat, stated in the RFC body: `ctx.llm.stream()` itself runs a further `llm/stream` waterfall before the adapter, so the guarantee is "authoritative within the agent request pipeline," not an absolute wire invariant; if a hard invariant is ever required, a defensive `llm/stream` assertion with a spy adapter covers it in tests. - -**3c. The single tool — `run_code`.** Registered normally in `ctx.tools` with one parameter `{ code: string (required) }`. Because it is an ordinary tool, the unchanged loop dispatches it through the normal path — this is the crux of "zero loop changes." Its `execute(args, exec)`: - -1. Builds the SDK bindings. For each real tool, an async `invoke(callArgs)` that **checks `exec.signal?.aborted` (throwing if set) before and after** calling `ctx.tools.execute({ callId: , name, arguments: callArgs, agent: exec.agent, signal: exec.signal })`, then maps the resulting `ContentBlock[]` to a simplified `{ output, isError }` (text blocks for the MVP), and emits an observability event. The explicit abort check matters because `ctx.tools.execute()` *catches* thrown tool errors and converts them to `isError` results — without the check, an aborted sub-call would look like ordinary error data and the program would keep running instead of stopping. Sub-dispatch still flows through the `tools/pre-execute`/`tools/post-execute` waterfalls, so permission/sandbox/hook plugins apply to code-mode calls exactly as to native ones. -2. Calls `ctx.codeRuntime.run({ code: args.code, sdk: bindings, signal: exec.signal })`. -3. Surfaces the outcome. A *successful* run returns `[{ type: 'text', text: }]`. A *runtime-error* result cannot be reported by returning content, because a normal `ToolDefinition.execute()` returns only `Promise` and `ToolRegistry.execute()` hardcodes `isError: false` on any successful return — `isError: true` arises only from the registry's catch path. So on an error result the tool **throws a `CodeRunError extends HarnessError`** (`HarnessError` is exported from `dsh-llm`; the registry catch turns any throw into `isError: true` with the message as text, and a `HarnessError` additionally carries structured `{ name, code }`). An alternative — registering `run_code` handling as a `tools/execute` listener that returns a full `ToolExecutionResult` and can set `isError` directly — is noted; the throw is simpler and preferred. - -**3d. Result discipline — what the model receives.** The model gets back **only the captured console output and/or the program's return value** (the model chooses which to surface). Intermediate sub-call results are **never** returned to the model. This is the core context-saving benefit: the agent curates its own output, exactly as a script's stdout curates a pipeline's intermediate state. - -**Sub-call CallIds.** Real tool calls dispatched from inside `run_code` need ids, but `CallId` is normally provider-issued (a branded string for correlating a call with its result — only brand-wrapped via `CallId()`, with no generator and no documented session-global-uniqueness guarantee). The plugin mints deterministic sub-ids scoped to the parent: `` `${exec.callId}:code:${n}` `` with a per-run counter `n`. These are unique within one `run_code` run (assuming the parent `callId` is unique, which the provider guarantees per turn); the `code/dispatch` event additionally carries the session log's `seq` so the UI and persistence can order and disambiguate globally without relying on the id alone. `ToolExecution.agent` is optional; the normal loop always supplies it (and with it `exec.agent.session`, the log `code/dispatch` appends to). A `run_code` execution arriving without `exec.agent` still runs (sub-calls propagate `agent: undefined`, exactly as the loop's own contract allows) but **skips session-log observability** — with no session to append to, those direct runs are simply not logged. - -**Observability without context cost.** Each sub-dispatch emits a session event **declared by the `dsh-code-mode` plugin itself** via `SessionEventMap` declaration merging (the map is merge-extensible precisely so plugins can add events without touching `dsh-session`). Shape: `code/dispatch` with `{ parentCallId, subCallId, name, arguments (or redacted), isError, summary }`, ordered by the session log's own `seq`. `deriveMessages()` does **not** translate it into a model message — an unknown event type falls through its `default`, per the merge-extensible-union convention — so the UI and persistence ([session persistence](../../implemented/architecture/2026-06-14-session-persistence.md)) can render every sub-call while the model's context only ever receives the single `run_code` tool-result. Because the event lives in the plugin, this adds no core change. - -**SDK codegen.** A pure `jsonSchemaToTs(schema)` in `code-mode` maps the JSON-schema subset the `defineTool` DSL produces (object/string/number/boolean/array, `properties`, `required[]`, `enum` → string-literal union, nested objects, array `items`) to a TS type literal. It is **total**: any unsupported construct (`$ref`, `oneOf`/`anyOf`, `integer`, `null`, `additionalProperties`, or any raw MCP shape it does not recognize) degrades to `unknown` without throwing — it never crashes codegen. Typing is best-effort, not a guarantee, because MCP tools accept arbitrary JSON Schema and `ToolSchema.parameters` is typed only as `Record`. Because `ToolSchema.name` is an arbitrary string (not necessarily a valid TS identifier), the SDK is generated as a **namespace with quoted access** (e.g. `tools["some-mcp-tool"](args)`) plus safe camelCase aliases where the name is a clean identifier; alias collisions and TS reserved words fall back to quoted-only access (no duplicate alias emitted). This mirrors Cloudflare's `sanitizeToolName`. `run_code` itself is filtered out of the SDK. The MVP surfaces text content only; image and other block types in sub-results are deferred (noted as a limitation). - -**Concurrency — serialized by default (the binding must enforce it).** The SDK functions are async, so a model writing `await Promise.all([tools.a(...), tools.b(...)])` would *start both* immediately, and each would call `ctx.tools.execute` right away — i.e. the binding shape makes concurrent dispatch the **default**, not an opt-in. Because the tool contract carries **no concurrency-safety metadata today** (parallel tool execution and a concurrency-safety hint are an open TODO in both `dsh-tools` and [docs/architecture.md](../../../architecture.md): "phase 1 executes tool calls sequentially"), concurrent dispatch through a not-yet-hardened tool may race. So a prose "may serialize" is not sufficient. **Decision: the MVP SDK bindings enforce serialization** — each `run_code` invocation owns a per-run dispatch queue, and every `invoke()` chains onto it (`tail = tail.then(() => ctx.tools.execute(...))`), so even `Promise.all` over SDK calls executes them one at a time in submission order. This is a hard acceptance criterion, with a test that issues `Promise.all([...])` from a program and asserts the underlying `ctx.tools.execute` calls did **not** overlap (e.g. a probe tool records enter/exit and the test asserts no interleaving). The `.d.ts` may *describe* the model-visible functions as async (they are), but the implementation guarantees serial execution. Lifting serialization is deferred: only once a tool can declare itself read-only / concurrency-safe does the binding allow those specific tools to overlap. The same per-run queue is where the before/after abort checks (§3c) live, so an aborted run drains no further queued dispatches. - -**Tool visibility tiers (design intentionally skipped).** A natural extension is to mark each tool with a *visibility tier*: some tools "direct-call eligible" (still offered as native wire tools alongside `run_code`), some "code-mode only" (reachable solely from within a `run_code` program, never on the wire), and the default "both." This would let a deployment keep a few high-frequency or approval-gated tools as direct calls while routing the long tail through Code Mode, or hide composition-only primitives from the native surface entirely. This RFC notes the possibility but **intentionally skips the detailed design** — the per-tool metadata, how it interacts with the `agent/request` enforcement in 3b, and the presentation split in 3a are left to a follow-up. The MVP is the simple two-state model: Code Mode on (everything via `run_code`) or off (everything native). - -**Optionality / toggle.** Loading the `code-mode` plugin enables Code Mode for that context; not loading it leaves today's native tool-calling untouched. The two are mutually exclusive within one ctx, because Code Mode rewrites the wire tool list down to `[run_code]`. Per-agent selection via ctx forks, and the visibility tiers above, are future work; the MVP toggle is plugin presence. - -## Alternatives considered - -**Result elision / summarization over native tool-calling (the narrower route).** The Problem has two halves — context bloat (every intermediate `tool-result` re-enters context) and serial composition (one tool call per round-trip). The context-bloat half can be addressed *without* any code-execution runtime: keep provider tool-calling exactly as it is, and add a plugin on the `agent/request` waterfall (or a compaction pass akin to [the session-persistence work](../../implemented/architecture/2026-06-14-session-persistence.md)) that elides or summarizes older `tool-result` blocks before they re-enter the model's context — drop them past a window, replace large payloads with a digest, or keep only the blocks the model still references. This is strictly less invasive than Code Mode: no new runtime seam, no model-written programs, no new safety surface. It is the right tool if context growth is the only pain. - -It is insufficient for the **composition / round-trip** half, which is the decisive reason this RFC does not stop there. Elision still pays one model round-trip per tool call: a loop over N items is N turns, a branch on an intermediate value is a turn to fetch then a turn to act, and post-processing (filter, join, reduce) either happens in the model's head over full payloads or not at all. Code Mode collapses all of that into one program — the loop, the branch, the join run in the runtime, and only the curated result returns. Elision also cannot express fan-out or data-dependent control flow; it only shrinks what comes back. So the two are complementary, not competing: elision could even layer *under* Code Mode for the residual native-tool paths. The RFC chooses Code Mode because the round-trip/composition cost is the larger structural limit, and accepts the new code-execution surface as the price — which is exactly why the execution substrate is gated behind the enforceable safety guard (§2) and the hardened backend is a hard prerequisite for untrusted use. - -**Why not change the loop to dispatch native tool calls in parallel instead?** That is the other obvious answer to the round-trip cost, and it remains valid future work (it is the open `dsh-tools`/architecture.md TODO). But it is a core-loop change requiring the same concurrency-safety metadata Code Mode defers, and it still does not give the model *composition* (branch/loop/post-process between calls) — only parallelism of independent calls the model already decided to make in one step. Code Mode delivers composition with zero core change; parallel native dispatch and Code Mode can coexist later. - -## Plan - -1. Scaffold the interface package `packages/code-runtime/` per [the cookbook](../../../cookbook/adding-a-package.md): abstract `CodeRuntime extends Service` (`super(ctx, 'codeRuntime')`) with a readonly `safe: boolean`, the `declare module 'cordis'` ctx key, the `CodeRunRequest`/`CodeRunResult`/`SdkBinding` vocabulary, method contracts documented in JSDoc (what `run` captures, abort semantics, that an error is a result field not a throw, what `safe` means). HMR-safety test (dispose the contributing fiber, assert `ctx.codeRuntime` is gone). -2. Scaffold the implementation package `packages/code-runtime-vm/`: the node:vm stub — `safe = false`, a constructor that **throws unless given `{ unsafe: true }`**, transpile/type-erase, async-IIFE wrap, capturing `console`, SDK globals, return-value/logs/error capture, output cap, signal-tied timeout. Tests for output capture, return value, error-as-field, abort, the constructor refusal without `unsafe`, and a README documenting the "not a sandbox, trusted-only" caveat prominently. -3. Scaffold the consumer plugin `packages/code-mode/`: `jsonSchemaToTs` codegen with namespace/quoted-access + alias handling (unit tests, including non-identifier MCP names and unsupported-shape → `unknown`); the registered lazy `ctx.systemPrompt.section()` carrying the SDK `.d.ts`; the `agent/request` listener (`prepend: true`) collapsing `request.tools` to `[run_code]` after `await next()`; the **unsafe-runtime gate** (refuse to register `run_code` when `ctx.codeRuntime.safe === false` unless `allowUnsafeRuntime` is set); the `run_code` tool with the dispatch bridge (per-run serialization queue, deterministic sub-call ids, before/after abort checks, `CodeRunError` on error results); and the `code/dispatch` event declared here via `SessionEventMap` merge. Declare `inject = ['tools', 'systemPrompt', 'codeRuntime']`. -4. Tests: HMR-safety (dispose removes the tool, the section, and the listener); a waterfall test that the wire tool list is exactly `[run_code]` (spy adapter, asserting via `agent/request` and optionally `llm/stream`); an integration test that a program calling two tools returns only its printed/returned output (verify the world, not the self-report); a **serialization test** that `Promise.all([...])` over SDK calls does not overlap the underlying `ctx.tools.execute` invocations (a probe tool records enter/exit; assert no interleaving); `deriveMessages()` ignores `code/dispatch`; abort mid-program stops further dispatches; `CodeRunError` surfaces as `isError: true`; and the **unsafe-runtime refusal test** (§3, the VM-guard): with the unsafe flag unset, a non-mock agent's `run_code` is refused; with it set, the program runs. -5. Wire an example: `examples/coding-agent-code-mode` (or a config flag on the existing example) loading the trio. Running it against the node:vm stub requires both opt-ins (`VmCodeRuntime({ unsafe: true })` and `code-mode`'s `allowUnsafeRuntime`); the example sets them explicitly and comments why, or uses a mock model — a real model never reaches the unsandboxed stub without those deliberate flags. Add a `pnpm run demo:*` entry. -6. Docs: update [docs/architecture.md](../../../architecture.md) (a `ctx.codeRuntime` row in the service map, a Code Mode note under the tool pipeline / capability seams sections); add a [cookbook](../../../cookbook) note on writing a `CodeRuntime` backend; and **file the follow-up RFC for the hardened execution substrate** (the isolate/sandboxed-process design, the additional-language backends sketched in §1 — AssemblyScript/WASM, Python — with their per-language SDK generators, plus the tool-visibility-tier design skipped here). On landing, move this file to `implemented/` and update its row in [the RFC index](../../README.md). - -## Acceptance criteria - -- The three packages exist and pass their suites: `dsh-code-runtime` (the abstract seam), `dsh-code-runtime-vm` (the reference stub whose constructor throws without `{ unsafe: true }`), and `dsh-code-mode` (SDK codegen, the lazy prompt section, the `agent/request` collapse, the `run_code` tool). -- The wire tool list is exactly `[run_code]` under the plugin (spy-adapter test); the generated SDK covers every registered tool, with non-identifier names reachable via quoted access. -- A program calling two tools returns only its curated output; `code/dispatch` events land in the session log and never enter derived history. -- `Promise.all` over SDK calls does not overlap the underlying `ctx.tools.execute` invocations (the per-run serialization queue holds); an abort stops further dispatches. -- With `allowUnsafeRuntime` unset over an unsafe runtime, `run_code` is not registered and the wire tool list is unchanged from native. - -## Risks - -node:vm is not a sandbox. This is the single biggest caveat. Withholding `require`/`process` is not a boundary; the MVP runs at harness trust only; the hardened substrate is a hard prerequisite before any untrusted use and is the explicit subject of a follow-up RFC. The guard is enforceable, not just documented: the runtime exposes `safe: boolean`, the VM stub throws unless constructed with `{ unsafe: true }`, and `code-mode` refuses to register `run_code` over an unsafe runtime unless separately acknowledged (`allowUnsafeRuntime`) — production misuse requires two deliberate, greppable flags, and the refusal path is tested. - -Wrong seam would leak tools. If the wire tool list were enforced only in `system-prompt/assemble`, a later `agent/request` listener could re-add tools. Mitigation: enforce `request.tools = [run_code]` in the `agent/request` waterfall (the authoritative seam, run last before `llm.stream()`) with `prepend: true`, and assert exactly one wire tool in tests. The residual `llm/stream` caveat is documented, not hidden. - -Concurrency before the contract supports it. The binding shape makes concurrent dispatch the default, and the tool contract has no concurrency-safety metadata yet, so unguarded `Promise.all` over SDK calls could race a not-yet-hardened tool. Mitigation: the MVP bindings enforce a per-run serialization queue (every `invoke` chains onto the previous), with a test asserting `Promise.all` from a program does not overlap the underlying `ctx.tools.execute` calls. Per-tool parallelism is unlocked only once a tool can declare itself concurrency-safe. - -Two presentation modes to keep coherent. A tool added later must work in both native and Code Mode. Mitigation: both the codegen thunk and the `agent/request` listener read `ctx.tools.schemas()`, so coverage is automatic; a test asserts every registered schema produces valid `.d.ts`, including non-identifier MCP names via quoted access. - -Type-erased runtime is not type-checked. The model can write code that type-checks against the advisory `.d.ts` but throws at runtime, and MCP-schema typing is best-effort. Mitigation: errors are captured as `CodeRunResult.error` and surfaced so the model can self-correct; the `.d.ts` is explicitly advisory. - -Lost observability of sub-calls. Routing everything through one `run_code` result hides the individual calls from the model — and could hide them from operators too. Mitigation: the plugin-declared `code/dispatch` event keeps every sub-call in the session log and UI without polluting model context. - -Abort granularity. node:vm cannot reliably interrupt hot synchronous code, and `ctx.tools.execute()` converts thrown aborts into `isError` data. Mitigation: the SDK bindings check `signal.aborted` and throw before/after each dispatch so an aborted sub-call stops the program; the vm stub wraps the run in a signal-tied timeout; the hardened substrate addresses the hot-loop case. - -Unsafe example wiring. A demo running a real model through the node:vm stub would hand model output ambient authority. Mitigation: examples are mock-model or explicitly marked unsafe; `code-runtime-vm` is labeled reference/test-only. - -Non-text sub-results dropped in the MVP. Image and other block types from sub-calls are not surfaced into the program yet. Mitigation: noted as a known limitation; block-type handling deferred. From 00ee92b278d8c35555cd8a5492eab88ad43c0179 Mon Sep 17 00:00:00 2001 From: Tianyi Cui <53024+tianyicui@users.noreply.github.com> Date: Wed, 8 Jul 2026 00:58:26 +0800 Subject: [PATCH 10/26] =?UTF-8?q?docs:=20record=20the=20codeRuntime=20cons?= =?UTF-8?q?umption=20idiom=20=E2=80=94=20cordis=20has=20no=20optional=20in?= =?UTF-8?q?ject?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Own adversarial pass finding: a static inject on the registry would gate ctx.tools (and every tool plugin) on a code runtime existing even under mode 'native'. The RFC now names the sanctioned pattern: soft ctx.get('codeRuntime') at use time (the agent-loop sessionPersistence precedent) with absence failing loud in the provider thunk. --- docs/rfc/proposed/feature/2026-06-15-code-mode.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/rfc/proposed/feature/2026-06-15-code-mode.md b/docs/rfc/proposed/feature/2026-06-15-code-mode.md index d868c41faf..1ca5b52057 100644 --- a/docs/rfc/proposed/feature/2026-06-15-code-mode.md +++ b/docs/rfc/proposed/feature/2026-06-15-code-mode.md @@ -59,7 +59,7 @@ Each sub-dispatch appends one session event, declared by `dsh-tools` via `Sessio - `CodeRunFailure = { kind: 'exception' | 'timeout' | 'abort' | 'worker-exit'; message: string }` — orthogonal outcomes reported independently per [defensive patterns](../../../defensive-patterns.md); a timed-out run is not an exception, an abort is not a timeout. - Two readonly backend descriptors, informational not gating: `language` (what the program must be written in — `'typescript'` for the shipped backend; a Python backend would say so, and pair with its own SDK generator on the presentation side) and `isolation` (`'worker-thread'` for the shipped backend; `'process'`, `'container'`, … for future ones). `dsh-tools` requires `language === 'typescript'` in the MVP — its codegen emits TS — and fails the assembly loudly otherwise, the same misconfiguration idiom as `toolOrder` violations (as when `mode` is non-native with no `ctx.codeRuntime` loaded at all). -Per explicit-over-implicit at seams, the request spells out everything the runtime acts on; defaulting (timeouts, caps) is the implementation's validated config, never a hidden `??` inside `run()`. The seam split is justified by real planned divergence on both axes — substrate (worker now; container/microVM later) and language (the Python/AssemblyScript direction sketched in the earlier draft survives as future work) — not by speculation: `dsh-tools` consumes the interface today and tests against a trivial in-repo fake, exactly the interface/implementation/consumer shape of the bash template. +Per explicit-over-implicit at seams, the request spells out everything the runtime acts on; defaulting (timeouts, caps) is the implementation's validated config, never a hidden `??` inside `run()`. Consumption uses the loop's established optional-backend idiom: cordis has no optional injection — every `inject` entry gates activation — so a static `inject` on the registry would hold `ctx.tools` (and every tool plugin behind it) hostage to a code runtime existing even under `mode: 'native'`; instead the registry reads `ctx.get('codeRuntime')` at use time, exactly as `agent-loop` consumes `sessionPersistence`, with absence failing loud in the provider thunk as above. The seam split is justified by real planned divergence on both axes — substrate (worker now; container/microVM later) and language (the Python/AssemblyScript direction sketched in the earlier draft survives as future work) — not by speculation: `dsh-tools` consumes the interface today and tests against a trivial in-repo fake, exactly the interface/implementation/consumer shape of the bash template. ### The worker-thread runtime From e90103ac0533c3bb1f7950668abc84aa40c0d293 Mon Sep 17 00:00:00 2001 From: Tianyi Cui <53024+tianyicui@users.noreply.github.com> Date: Wed, 8 Jul 2026 01:02:37 +0800 Subject: [PATCH 11/26] docs: name the persistence-catalog gate for the tool/code-dispatch event MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Research finding: a SessionEventMap member is a log event — JSDoc prose required, @mode is a hard error there, and docs/persistence-catalog.md must be regenerated (todo/write is the log-only precedent). PR4's plan now names both. --- docs/rfc/proposed/feature/2026-06-15-code-mode.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/docs/rfc/proposed/feature/2026-06-15-code-mode.md b/docs/rfc/proposed/feature/2026-06-15-code-mode.md index 1ca5b52057..e4f4c5a32d 100644 --- a/docs/rfc/proposed/feature/2026-06-15-code-mode.md +++ b/docs/rfc/proposed/feature/2026-06-15-code-mode.md @@ -46,7 +46,7 @@ Under `'code'` and `'both'` the registry registers `run_code` in itself as an or ### Observability: `tool/code-dispatch` -Each sub-dispatch appends one session event, declared by `dsh-tools` via `SessionEventMap` declaration merging (the map is merge-extensible for exactly this): `tool/code-dispatch` with `{ parentCallId, subCallId, name, arguments, isError, resultSummary }`. It is log-only — `deriveEventMessage()` ignores unknown event types by design, so sub-calls never re-enter model context — but persistence and UIs get every call. A `run_code` execution arriving without `exec.agent` (the loop always supplies it; direct programmatic calls may not) still runs and simply skips event logging, exactly as the `ToolExecution` contract allows. +Each sub-dispatch appends one session event, declared by `dsh-tools` via `SessionEventMap` declaration merging (the map is merge-extensible for exactly this; `todo/write` is the log-only precedent): `tool/code-dispatch` with `{ parentCallId, subCallId, name, arguments, isError, resultSummary }`. It is log-only — `deriveEventMessage()` ignores unknown event types by design, so sub-calls never re-enter model context — but persistence and UIs get every call. As a log event it carries JSDoc prose but **no `@mode` tag** (that vocabulary belongs to cordis bus events; the persistence-catalog generator hard-errors on one) and lands in the regenerated `docs/persistence-catalog.md`; appends happen inside `run_code`'s execution, so the turn-enclosure invariant is satisfied by construction. A `run_code` execution arriving without `exec.agent` (the loop always supplies it; direct programmatic calls may not) still runs and simply skips event logging, exactly as the `ToolExecution` contract allows. ### The code-runtime seam @@ -87,7 +87,7 @@ Four stacked PRs, each gates-green (`typecheck`, `lint`, `test:coverage`, `test: 1. **This RFC revision** (docs-only): the file rewritten as above (renamed `2026-06-15-code-mode.md`, same first-proposed date), regenerated RFC index. 2. **`dsh-code-runtime`** (interface package): the group `packages/code-runtime/`, abstract `CodeRuntime`, vocabulary types, ctx-key declaration, group README + package README + catalog updates. Unit tier: HMR safety (dispose removes `ctx.codeRuntime`), contract docs. This package has no behavior to snapshot or e2e; its coverage story is unit-only by design. 3. **`dsh-code-runtime-worker`**: the implementation above. Unit tier (real workers, no mocks — they are cheap and local): output/value capture, log source attribution, error kinds (exception incl. non-erasable syntax, timeout vs abort vs worker-exit under OOM), compute-time-only timeout (a slow binding does not expire the run), binding bridge hostility cases (unknown name, duplicate id, post-settlement message), structured-clone fallback, cap truncation, `env` emptiness verified from inside the program, dispose-awaits-exit. A real-load-path test runs the built package (`lib/`) so the worker entry resolves both unbuilt (tsx) and built — the published-bin guard from [docs/testing.md](../../../testing.md). -4. **Native code mode in `dsh-tools`** + the end-to-end surface: mode config, provider switch, `tools:sdk` section, `jsonSchemaToTs`, `run_code` + dispatch bridge + serialization queue, `tool/code-dispatch` event; [docs/architecture.md](../../../architecture.md) (tool-pipeline section + `ctx.codeRuntime` service row) and the [adding-a-tool cookbook](../../../cookbook/adding-a-tool.md) cross-reference updated in the same change; an `examples/` leaf + `demo:code` script wiring the worker runtime with `mode: 'code'`; move this RFC to `implemented/`. Coverage named per tier now, per the plan-time rule: **unit** — codegen table (DSL subset, quoted names, `unknown` degradation, determinism), provider contribution per mode, `toolOrder × mode` rejection, missing-runtime/wrong-language loud failures, serialization non-overlap (a probe tool records enter/exit under `Promise.all`), abort stops the queue, binding rejection on `isError`, `CodeRunFailedError` → structured `isError`, event payloads, `deriveMessages()` ignores the event, HMR safety (mode flip via config reload removes tool + section); **e2e (with-key, self-skips)** — a real model, `mode: 'code'`, a task requiring two tool calls and curation, asserting the wire tool list was exactly `run_code` and the transcript's dispatch events; **snapshot (keyless replay)** — goldens for a `run_code` turn in `'code'` and `'both'`, pinning the SDK section text, the collapsed header tools, dispatch events, and the result card. +4. **Native code mode in `dsh-tools`** + the end-to-end surface: mode config, provider switch, `tools:sdk` section, `jsonSchemaToTs`, `run_code` + dispatch bridge + serialization queue, `tool/code-dispatch` event (+ regenerated `docs/persistence-catalog.md`); [docs/architecture.md](../../../architecture.md) (tool-pipeline section + `ctx.codeRuntime` service row) and the [adding-a-tool cookbook](../../../cookbook/adding-a-tool.md) cross-reference updated in the same change; an `examples/` leaf + `demo:code` script wiring the worker runtime with `mode: 'code'`; move this RFC to `implemented/`. Coverage named per tier now, per the plan-time rule: **unit** — codegen table (DSL subset, quoted names, `unknown` degradation, determinism), provider contribution per mode, `toolOrder × mode` rejection, missing-runtime/wrong-language loud failures, serialization non-overlap (a probe tool records enter/exit under `Promise.all`), abort stops the queue, binding rejection on `isError`, `CodeRunFailedError` → structured `isError`, event payloads, `deriveMessages()` ignores the event, HMR safety (mode flip via config reload removes tool + section); **e2e (with-key, self-skips)** — a real model, `mode: 'code'`, a task requiring two tool calls and curation, asserting the wire tool list was exactly `run_code` and the transcript's dispatch events; **snapshot (keyless replay)** — goldens for a `run_code` turn in `'code'` and `'both'`, pinning the SDK section text, the collapsed header tools, dispatch events, and the result card. The four PRs land in order (each on the previous); per stacked-review practice, review fixes land on the PR that introduced them and merge down. From cb3246d2d875e5195902f6b2c01c9307bfbc48fa Mon Sep 17 00:00:00 2001 From: Tianyi Cui <53024+tianyicui@users.noreply.github.com> Date: Wed, 8 Jul 2026 01:26:33 +0800 Subject: [PATCH 12/26] docs: fix Codex round-1 findings on the Code Mode RFC (A1) Scope the wire-collapse guarantee honestly: systemPrompt.tools() is a public multi-provider API, so the mode governs the registry's contribution (the only shipped source); deliberate extra providers own what they add, and the shipped-configuration invariant is test-pinned. (A2) Replace pause-on-pending-RPC timeout with two independent budgets: computeMs metered by worker.performance.eventLoopUtilization() busy time (unfoolable by an un-awaited decoy dispatch; probe-verified) plus a never-pausing maxWallMs ceiling. (A3) Specify sub-call additionalContext as deliberately suppressed in the MVP (immediate inject would break call/result adjacency; the plural channel is named follow-up work). (B) Orphan-process caveat vs bash-local's group kill; null-prototype binding namespaces (__proto__/constructor names); per-PR doc artifacts (packages/README row, architecture service map in PR2, config/tool/ persistence catalogs per owning PR); engines range corrected to ^22.19.0 || >=24.0.0. --- .../proposed/feature/2026-06-15-code-mode.md | 24 ++++++++++--------- 1 file changed, 13 insertions(+), 11 deletions(-) diff --git a/docs/rfc/proposed/feature/2026-06-15-code-mode.md b/docs/rfc/proposed/feature/2026-06-15-code-mode.md index e4f4c5a32d..67f26ad688 100644 --- a/docs/rfc/proposed/feature/2026-06-15-code-mode.md +++ b/docs/rfc/proposed/feature/2026-06-15-code-mode.md @@ -24,7 +24,7 @@ Three decisions, each elaborated in its own section below: `ToolRegistry` gains a schemastery-validated config (`static Config`), its first: `mode: 'native' | 'code' | 'both'`, default `'native'`. A deployment flips it from `cordis.yml` (`tools: { mode: code }`) — no code edit, per the no-hardcoded-tunables convention. -**Wire tool list = the provider's contribution.** The registry already feeds the assembly through `ctx.systemPrompt.tools(() => this.schemas())`; the provider becomes mode-aware: `'native'` contributes all schemas (unchanged), `'code'` contributes only `run_code`'s schema, `'both'` contributes all schemas plus `run_code`. Because [`PromptAssembly.tools` is the single source the loop's request header snapshots](../../../../packages/core/system-prompt/src/index.ts), the collapse is automatically logged and reconstructable — model-visible ⟺ logged holds with zero new mechanism, and there is no seam left where another party could accidentally re-add tools (the `system-prompt/assemble` waterfall can still deliberately transform the assembly, as ever). +**Wire tool list = the registry's contribution.** The registry already feeds the assembly through `ctx.systemPrompt.tools(() => this.schemas())`; the provider becomes mode-aware: `'native'` contributes all schemas (unchanged), `'code'` contributes only `run_code`'s schema, `'both'` contributes all schemas plus `run_code`. Because [`PromptAssembly.tools` is the single source the loop's request header snapshots](../../../../packages/core/system-prompt/src/index.ts), the collapse is automatically logged and reconstructable — model-visible ⟺ logged holds with zero new mechanism. Scope of the guarantee, stated honestly: the mode governs the **registry's** contribution, and the registry is the only shipped schema source — but `systemPrompt.tools()` is a public multi-provider API and the `system-prompt/assemble` waterfall may transform the assembly, so a deployment that wires a second direct provider (or a mutating listener) owns what it adds, exactly as in native mode. Those are deliberate acts; what the design eliminates is the *accidental* leak the old draft worried about — a listener-ordering race around an after-the-fact collapse — and the shipped-configuration invariant (`'code'` ⇒ assembled tools exactly `[run_code]`) is pinned by tests and, like every request, by the logged header. **Interaction with `toolOrder`, stated up front:** a configured `systemPrompt.toolOrder` naming native tools rejects every assembly under `mode: 'code'` (those names are no longer contributed), by the existing fail-loud rule for unlisted names. This is correct behavior, not a bug: a deployment switching modes updates its order config or drops it. @@ -40,6 +40,8 @@ Under `'code'` and `'both'` the registry registers `run_code` in itself as an or 2. **Runs the program**: `ctx.codeRuntime.run({ program: args.code, bindings: [{ global: 'tools', functions }], signal: exec.signal })`. 3. **Surfaces the outcome**: a successful run returns one text block — the captured console/stdout output followed by the rendered return value (if any) — plus a `meta` payload (capped logs, dispatch count) for presentation. A run with `result.error` throws a `CodeRunFailedError extends HarnessError` (`code: 'CODE_RUN_FAILED'`, message = the error kind and text plus captured logs so the model can self-correct); the registry's existing catch turns it into a structured `isError` result. +**Sub-call `additionalContext` is suppressed, deliberately.** A `tools/post-execute` hook may attach `additionalContext` to a call; for loop-dispatched calls the loop buffers those and appends each as a `context/message` only after the step's `tool/result`s, preserving call/result adjacency. A sub-dispatch result's `additionalContext` has no such safe outlet from inside a running `run_code`: injecting immediately would land a `context/message` between the parent's `tool/call` and its `tool/result` (breaking the adjacency the buffering exists to protect), and `PostToolDecision.additionalContext` is singular where a program may produce many. The MVP therefore drops sub-call `additionalContext`, pinned by a test and stated in the hooks bridge's docs; the follow-up (a plural context channel or loop-level sub-dispatch buffering) is deferred until a real hook needs it through Code Mode. + **Concurrency: serialized, enforced by the binding.** The bindings are async, so a model writing `Promise.all([tools.a(…), tools.b(…)])` starts both immediately — concurrent dispatch would be the *default*, while the tool contract still carries no concurrency-safety metadata (the open parallel-execution TODO). Each `run_code` invocation therefore owns a dispatch queue and every binding call chains onto it, so even `Promise.all` executes the underlying `ctx.tools.execute()` calls one at a time in submission order; when the run settles, queued-but-unstarted dispatches are abandoned. Lifting this per-tool once tools can declare themselves concurrency-safe is deferred work, same as before. **Presentation.** `run_code`'s render intent is decided here per the [render-intent RFC](../../implemented/architecture/2026-07-02-tool-render-intent-union.md): `presentCall` → a `generic` card, `kind: 'execute'`, title `Run code`, `rawInput` = the program text; `presentResult` → a `generic` card whose content is the captured output (from `meta`). Not a `terminal` card: that card's semantics are "a shell command in a working directory", which a program is not. @@ -65,16 +67,16 @@ Per explicit-over-implicit at seams, the request spells out everything the runti `packages/code-runtime/code-runtime-worker/` — `@deepseek-ai/dsh-code-runtime-worker`. Per `run()`: -1. **Type-strip host-side** with Node's built-in `stripTypeScriptTypes` (`node:module`; the repo floor is node ≥ 24, and the API is position-preserving, so runtime error line numbers match the model's source). Strip-only mode rejects non-erasable syntax (`enum`, namespaces) — that rejection returns as `error.kind: 'exception'` with Node's message, the SDK instructions say "erasable TypeScript only", and the model self-corrects like any other program error. A syntax-level failure never spawns a worker. +1. **Type-strip host-side** with Node's built-in `stripTypeScriptTypes` (`node:module`; present across the repo's whole engines range, `^22.19.0 || >=24.0.0`, and position-preserving, so runtime error line numbers match the model's source). Strip-only mode rejects non-erasable syntax (`enum`, namespaces) — that rejection returns as `error.kind: 'exception'` with Node's message, the SDK instructions say "erasable TypeScript only", and the model self-corrects like any other program error. A syntax-level failure never spawns a worker. 2. **Spawn one fresh `Worker` per run** from the package's own bootstrap module: `env: {}` (truly empty — stronger than the scrubbed-env rule for spawned commands), `resourceLimits` from config, `stdout`/`stderr` captured into `logs` rather than inherited. No pooling and no cross-run state: a program's world dies with its worker, which keeps runs reconstructable from the log alone and makes state bleed unrepresentable. 3. **Execute** in the bootstrap: the stripped program becomes the body of an `AsyncFunction` whose parameters are the binding globals and a capturing `console` shim, so top-level `await` and `return` work and the program's completion value is the run's `value` (structured-cloneable values cross as-is; anything else is replaced by its `util.inspect` rendering, documented). -4. **Bridge bindings over the message port**: each binding function in the worker posts `{ id, global, name, args }` and awaits the reply; the host validates the name against the request's bindings, invokes, and replies `{ id, ok, value }` or `{ id, ok: false, message }` (a host-side binding rejection becomes a program-side rejection). Unknown names, duplicate ids, and post-settlement messages are rejected or ignored — the port protocol assumes a hostile peer, because the peer runs model code. -5. **Enforce caps**: the compute-time budget (`timeoutMs`) ticks only while no binding call is pending — it bounds runaway worker compute, not tool latency — and expiry, `signal` abort, and run completion all funnel into `worker.terminate()`, which ends hot synchronous loops too (measured; this was `node:vm`'s unfixable gap). Heap overflow surfaces as the worker's OOM exit → `error.kind: 'worker-exit'`. Log and value sizes are capped by config (`maxLogBytes`, `maxValueBytes`), truncation marked in-band. All caps are validated config fields with defaults (`timeoutMs: 60_000`, `maxLogBytes: 65_536`, `maxValueBytes: 32_768`, `maxOldGenerationSizeMb: 512`), changeable from `cordis.yml`. +4. **Bridge bindings over the message port**: each binding function in the worker posts `{ id, global, name, args }` and awaits the reply; the host validates the name against the request's bindings, invokes, and replies `{ id, ok, value }` or `{ id, ok: false, message }` (a host-side binding rejection becomes a program-side rejection). The worker-side namespace objects are built null-prototype via `defineProperty`, so a binding named `__proto__`, `constructor`, or `toString` is an ordinary own property, not a prototype collision. Unknown names, duplicate ids, and post-settlement messages are rejected or ignored — the port protocol assumes a hostile peer, because the peer runs model code. +5. **Enforce caps — two independent budgets, because the peer is hostile.** The compute budget (`computeMs`) meters the worker's *measured busy time* via `worker.performance.eventLoopUtilization()` polling — not host-side "is an RPC pending" bookkeeping, which a program defeats by firing an un-awaited call at a slow tool and then spinning hot while the host thinks it is waiting. Measured busy time cannot be gamed: a hot loop accrues it whether or not a dispatch is in flight, and a program genuinely awaiting a slow tool accrues none, so a long-running `bash` sub-call still does not kill an innocent run. The wall ceiling (`maxWallMs`) never pauses for anything and backstops what busy-time cannot see (a program awaiting a promise nobody will resolve). Budget expiry, `signal` abort, and run completion all funnel into `worker.terminate()`, which ends hot synchronous loops too (measured; this was `node:vm`'s unfixable gap); the failure reports which budget fired. Heap overflow surfaces as the worker's OOM exit → `error.kind: 'worker-exit'`. Log and value sizes are capped by config, truncation marked in-band. All caps are validated config fields with defaults (`computeMs: 60_000`, `maxWallMs: 600_000`, `maxLogBytes: 65_536`, `maxValueBytes: 32_768`, `maxOldGenerationSizeMb: 512`), changeable from `cordis.yml`. 6. **Dispose to quiescence**: the service's own disposal terminates in-flight workers and *awaits* their exits before resolving, per [defensive patterns](../../../defensive-patterns.md). ### Trust posture -The worker runtime is **containment, not a security boundary**, and the RFC says so without ceremony. Model code in the worker can reach Node globals — `fetch`, `process` (with an empty env) — so a deliberately adversarial program has ambient authority comparable to what the harness's own `bash` tool already grants every model turn: `dsh-bash-local` runs arbitrary model-written commands with the host filesystem, network, and a scrubbed-but-populated environment. Code Mode is gated where bash is gated — `tools/pre-execute`, where permission/sandbox plugins veto or approve the program before it runs — and adds containment bash does not have: empty env, heap caps, hard termination, a separate isolate. The earlier draft's two-flag unsafe ceremony (`{ unsafe: true }` constructor + `allowUnsafeRuntime`) existed for a `node:vm` stub with *no* containment and is dropped with it; demanding scarier flags for the better-contained executor than for bash would be posture theater. A deployment that needs a hard boundary (untrusted multi-tenant input) needs it for bash too; that is a future `isolation: 'container'` backend, and the `isolation` descriptor exists so such a deployment can tell backends apart. +The worker runtime is **containment, not a security boundary**, and the RFC says so without ceremony. Model code in the worker can reach Node globals — `fetch`, `process` (with an empty env), dynamic `import()` of built-ins — so a deliberately adversarial program has ambient authority comparable to what the harness's own `bash` tool already grants every model turn: `dsh-bash-local` runs arbitrary model-written commands with the host filesystem, network, and a scrubbed-but-populated environment. One asymmetry runs the other way and is stated plainly: `worker.terminate()` ends the thread, not OS processes a program may have spawned via `node:child_process` — weaker than `bash-local`'s process-group kill for direct children (equivalent for double-forked daemons, which survive both); the wall-clock ceiling bounds the worker itself, and orphan cleanup is the same deployment-level concern it already is for bash. Code Mode is gated where bash is gated — `tools/pre-execute`, where permission/sandbox plugins veto or approve the program before it runs — and adds containment bash does not have: empty env, heap caps, hard termination of the program itself, a separate isolate. The earlier draft's two-flag unsafe ceremony (`{ unsafe: true }` constructor + `allowUnsafeRuntime`) existed for a `node:vm` stub with *no* containment and is dropped with it; demanding scarier flags for the better-contained executor than for bash would be posture theater. A deployment that needs a hard boundary (untrusted multi-tenant input) needs it for bash too; that is a future `isolation: 'container'` backend, and the `isolation` descriptor exists so such a deployment can tell backends apart. ### What the model sees @@ -85,9 +87,9 @@ The `tools:sdk` section carries the `.d.ts` plus fixed instructions: the program Four stacked PRs, each gates-green (`typecheck`, `lint`, `test:coverage`, `test:snapshot`, `doc-sync`, `verify-module-graph`, `build`, `hygiene`) with docs updated in the same change: 1. **This RFC revision** (docs-only): the file rewritten as above (renamed `2026-06-15-code-mode.md`, same first-proposed date), regenerated RFC index. -2. **`dsh-code-runtime`** (interface package): the group `packages/code-runtime/`, abstract `CodeRuntime`, vocabulary types, ctx-key declaration, group README + package README + catalog updates. Unit tier: HMR safety (dispose removes `ctx.codeRuntime`), contract docs. This package has no behavior to snapshot or e2e; its coverage story is unit-only by design. -3. **`dsh-code-runtime-worker`**: the implementation above. Unit tier (real workers, no mocks — they are cheap and local): output/value capture, log source attribution, error kinds (exception incl. non-erasable syntax, timeout vs abort vs worker-exit under OOM), compute-time-only timeout (a slow binding does not expire the run), binding bridge hostility cases (unknown name, duplicate id, post-settlement message), structured-clone fallback, cap truncation, `env` emptiness verified from inside the program, dispose-awaits-exit. A real-load-path test runs the built package (`lib/`) so the worker entry resolves both unbuilt (tsx) and built — the published-bin guard from [docs/testing.md](../../../testing.md). -4. **Native code mode in `dsh-tools`** + the end-to-end surface: mode config, provider switch, `tools:sdk` section, `jsonSchemaToTs`, `run_code` + dispatch bridge + serialization queue, `tool/code-dispatch` event (+ regenerated `docs/persistence-catalog.md`); [docs/architecture.md](../../../architecture.md) (tool-pipeline section + `ctx.codeRuntime` service row) and the [adding-a-tool cookbook](../../../cookbook/adding-a-tool.md) cross-reference updated in the same change; an `examples/` leaf + `demo:code` script wiring the worker runtime with `mode: 'code'`; move this RFC to `implemented/`. Coverage named per tier now, per the plan-time rule: **unit** — codegen table (DSL subset, quoted names, `unknown` degradation, determinism), provider contribution per mode, `toolOrder × mode` rejection, missing-runtime/wrong-language loud failures, serialization non-overlap (a probe tool records enter/exit under `Promise.all`), abort stops the queue, binding rejection on `isError`, `CodeRunFailedError` → structured `isError`, event payloads, `deriveMessages()` ignores the event, HMR safety (mode flip via config reload removes tool + section); **e2e (with-key, self-skips)** — a real model, `mode: 'code'`, a task requiring two tool calls and curation, asserting the wire tool list was exactly `run_code` and the transcript's dispatch events; **snapshot (keyless replay)** — goldens for a `run_code` turn in `'code'` and `'both'`, pinning the SDK section text, the collapsed header tools, dispatch events, and the result card. +2. **`dsh-code-runtime`** (interface package): the group `packages/code-runtime/`, abstract `CodeRuntime`, vocabulary types, ctx-key declaration; docs in the same change — group README + package README, the `packages/README.md` group table row, the `ctx.codeRuntime` row in [docs/architecture.md](../../../architecture.md)'s service map, and the regenerated cordis catalog (the new service class). Unit tier: HMR safety (dispose removes `ctx.codeRuntime`), contract docs. This package has no behavior to snapshot or e2e; its coverage story is unit-only by design. +3. **`dsh-code-runtime-worker`**: the implementation above, plus the regenerated config catalog (its `Config`). Unit tier (real workers, no mocks — they are cheap and local): output/value capture, log source attribution, error kinds (exception incl. non-erasable syntax, abort, worker-exit under OOM), the two budgets from both sides (a hot loop with an un-awaited pending dispatch still dies at `computeMs` busy time; a program idling on a slow binding outlives `computeMs` and dies only at `maxWallMs`), binding bridge hostility cases (unknown name, duplicate id, post-settlement message, `__proto__`/`constructor`/`toString` binding names), structured-clone fallback, cap truncation, `env` emptiness verified from inside the program, dispose-awaits-exit. A real-load-path test runs the built package (`lib/`) so the worker entry resolves both unbuilt (tsx) and built — the published-bin guard from [docs/testing.md](../../../testing.md). +4. **Native code mode in `dsh-tools`** + the end-to-end surface: mode config, provider switch, `tools:sdk` section, `jsonSchemaToTs`, `run_code` + dispatch bridge + serialization queue, `tool/code-dispatch` event; regenerated tool, config, and persistence catalogs; [docs/architecture.md](../../../architecture.md) (tool-pipeline prose) and the [adding-a-tool cookbook](../../../cookbook/adding-a-tool.md) cross-reference updated in the same change; an `examples/` leaf + `demo:code` script wiring the worker runtime with `mode: 'code'`; move this RFC to `implemented/`. Coverage named per tier now, per the plan-time rule: **unit** — codegen table (DSL subset, quoted names, `unknown` degradation, determinism), provider contribution per mode, `toolOrder × mode` rejection, missing-runtime/wrong-language loud failures, serialization non-overlap (a probe tool records enter/exit under `Promise.all`), abort stops the queue, binding rejection on `isError`, `CodeRunFailedError` → structured `isError`, event payloads, `deriveMessages()` ignores the event, HMR safety (mode flip via config reload removes tool + section); **e2e (with-key, self-skips)** — a real model, `mode: 'code'`, a task requiring two tool calls and curation, asserting the wire tool list was exactly `run_code` and the transcript's dispatch events; **snapshot (keyless replay)** — goldens for a `run_code` turn in `'code'` and `'both'`, pinning the SDK section text, the collapsed header tools, dispatch events, and the result card. The four PRs land in order (each on the previous); per stacked-review practice, review fixes land on the PR that introduced them and merge down. @@ -116,7 +118,7 @@ The four PRs land in order (each on the previous); per stacked-review practice, - The generated `.d.ts` covers every registered tool except `run_code`, non-identifier names via quoted keys, unsupported schema constructs as `unknown`, without codegen ever throwing. - A program calling two tools returns only its curated output; each sub-call appears as a `tool/code-dispatch` event ordered by log `seq`, flows through `tools/pre-execute`/`post-execute` (a deny reaches the program as a binding rejection), and never enters derived messages. - `Promise.all` over three bindings produces non-overlapping `ctx.tools.execute()` intervals (probe-tool assertion); aborting mid-program stops the worker and dispatches nothing further. -- Worker runtime: a hot `for(;;){}` run ends at the compute-time budget with `error.kind: 'timeout'`; a pending binding call does not consume that budget; OOM under `resourceLimits` yields `kind: 'worker-exit'` with the host process healthy; `process.env` inside a program is empty; non-erasable syntax yields `kind: 'exception'` without a worker spawn; disposal awaits worker exit. +- Worker runtime: a hot `for(;;){}` run ends at the `computeMs` busy-time budget with `error.kind: 'timeout'` — including when the program fired an un-awaited binding call first (the pending-RPC decoy); a program idling on a slow binding does not consume `computeMs` and is bounded only by `maxWallMs`; OOM under `resourceLimits` yields `kind: 'worker-exit'` with the host process healthy; `process.env` inside a program is empty; non-erasable syntax yields `kind: 'exception'` without a worker spawn; disposal awaits worker exit. - Misconfiguration is loud before any model request: non-native mode with no `ctx.codeRuntime`, a runtime whose `language ≠ 'typescript'`, and `toolOrder` naming a non-contributed tool all reject the assembly with actionable messages. - The demo runs against the real API via `demo:code`; the snapshot goldens replay keylessly; all repo gates pass on every PR of the stack. @@ -124,7 +126,7 @@ The four PRs land in order (each on the previous); per stacked-review practice, **The worker is not a hard security boundary.** Deliberate and documented (§Trust posture): posture equals the existing bash tool, containment exceeds it, gating uses the same seams. Deployments needing more need a future `isolation: 'container'` backend — tracked as the seam's designed extension, not a TODO on this design. -**`stripTypeScriptTypes` is marked experimental.** It is also what Node itself runs `.ts` files with on this repo's node floor. Mitigations: the runtime's unit suite pins the behaviors relied on (position preservation, erasable-only rejection message shape loosely), the call sits behind one private function, and `amaro`/`sucrase` are drop-in replacements if the API shifts. The erasable-only subset is a model-facing contract line, and the error path is a working feedback loop, not a dead end. +**`stripTypeScriptTypes` is marked experimental.** It is the same engine (amaro/swc) behind Node's own native `.ts` execution, exposed as an API across this repo's whole engines range. Mitigations: the runtime's unit suite pins the behaviors relied on (position preservation, erasable-only rejection message shape loosely), the call sits behind one private function, and `amaro`/`sucrase` are drop-in replacements if the API shifts. The erasable-only subset is a model-facing contract line, and the error path is a working feedback loop, not a dead end. **Prompt cost of the SDK, especially under `'both'`.** The `.d.ts` can rival the native schemas it complements; `'both'` carries two representations. Prefix stability + provider caching amortize per-session cost; the mode is per-deployment; the RFC makes no unconditional-savings claim. Measured guidance (when to prefer which mode) is explicitly post-ship learning. @@ -134,4 +136,4 @@ The four PRs land in order (each on the previous); per stacked-review practice, **Serialized-only sub-dispatch.** `Promise.all` gains no wall-clock parallelism yet, only fewer round-trips; models may over-expect. The instructions state it; lifting it is tied to the same concurrency-safety metadata the native parallel-dispatch TODO needs. -**Timer-pause bookkeeping.** Compute-time-only budgeting adds a small state machine (pause on RPC out, resume on reply) whose bugs would misattribute time. It is unit-tested from both sides (slow binding ≠ timeout; hot loop = timeout) and is still simpler than the alternative — one budget that spuriously kills programs for waiting on a slow tool. +**Budget metering reads the event loop, not a flag.** Busy-time polling (`eventLoopUtilization()`) is coarser than an exact CPU meter — a budget expires up to one poll interval late — and its correctness claim ("a pending dispatch cannot pause it") is load-bearing against a hostile program. Both sides are unit-tested (hot loop with a pending decoy dispatch dies at `computeMs`; idle-on-slow-binding survives to `maxWallMs`), and the poll interval is an internal constant, not config — nothing a deployment could mis-tune into a bypass. From 96de50cb4835c24fe48967db9a05277f8b98bdc0 Mon Sep 17 00:00:00 2001 From: Tianyi Cui <53024+tianyicui@users.noreply.github.com> Date: Wed, 8 Jul 2026 01:39:10 +0800 Subject: [PATCH 13/26] docs: add the run-local abort/settlement contract (Codex round 2) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Budget expiry terminated the worker but nothing cancelled an in-flight host-side sub-dispatch, and a late dispatch could append events after run_code returned. The bridge now owns a run-scoped AbortController (follows exec.signal; fired on any run settlement), sub-dispatches get the run signal, and run_code returns only after the dispatch queue drains — no post-settlement appends, per dispose-to-quiescence. --- docs/rfc/proposed/feature/2026-06-15-code-mode.md | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/docs/rfc/proposed/feature/2026-06-15-code-mode.md b/docs/rfc/proposed/feature/2026-06-15-code-mode.md index 67f26ad688..30f4d71319 100644 --- a/docs/rfc/proposed/feature/2026-06-15-code-mode.md +++ b/docs/rfc/proposed/feature/2026-06-15-code-mode.md @@ -36,9 +36,9 @@ Three decisions, each elaborated in its own section below: Under `'code'` and `'both'` the registry registers `run_code` in itself as an ordinary tool — one required parameter `{ code: string }` — so the unchanged loop dispatches it through the normal pipeline and `tools/pre-execute` / `tools/post-execute` gate it like any other call (a permission plugin can inspect the program text before it runs). Its `execute(args, exec)`: -1. **Builds the bindings**: for every registered tool except `run_code`, an async function that (a) checks `exec.signal?.aborted` before and after (throwing stops the program — necessary because `ctx.tools.execute()` converts errors to `isError` data), (b) awaits its turn on the **per-run serialization queue** (below), (c) calls `this.execute({ callId, name, arguments, agent: exec.agent, signal: exec.signal })` with a deterministic sub-id `` CallId(`${exec.callId}:code:${n}`) ``, (d) appends a `tool/code-dispatch` session event, and (e) maps the result: success → the text-block contents joined as a `string` (non-text blocks become placeholders, an MVP limitation), `isError` → **the binding rejects** with an `Error` carrying the result text. Rejection is the deliberate model-facing contract — real code signals failure by throwing, `try/catch` and `Promise.all` short-circuiting behave as every model has seen them behave — where the old draft's `{ output, isError }` envelope made error handling a bespoke convention. +1. **Builds the bindings**: the bridge owns a **run-scoped `AbortController`** whose signal follows `exec.signal` (an outer cancel propagates in) and which the bridge itself fires the moment the run settles for any reason — completion, program exception, `computeMs`/`maxWallMs` expiry, worker exit. For every registered tool except `run_code`, the binding is an async function that (a) checks the run signal before and after (throwing stops the program — necessary because `ctx.tools.execute()` converts errors to `isError` data), (b) awaits its turn on the **per-run serialization queue** (below), (c) calls `this.execute({ callId, name, arguments, agent: exec.agent, signal: runSignal })` with a deterministic sub-id `` CallId(`${exec.callId}:code:${n}`) `` — the run signal, not the bare outer one, so a budget expiry aborts an in-flight sub-tool (`bash-local` kills on its spec signal) instead of orphaning it, (d) appends a `tool/code-dispatch` session event, and (e) maps the result: success → the text-block contents joined as a `string` (non-text blocks become placeholders, an MVP limitation), `isError` → **the binding rejects** with an `Error` carrying the result text. Rejection is the deliberate model-facing contract — real code signals failure by throwing, `try/catch` and `Promise.all` short-circuiting behave as every model has seen them behave — where the old draft's `{ output, isError }` envelope made error handling a bespoke convention. 2. **Runs the program**: `ctx.codeRuntime.run({ program: args.code, bindings: [{ global: 'tools', functions }], signal: exec.signal })`. -3. **Surfaces the outcome**: a successful run returns one text block — the captured console/stdout output followed by the rendered return value (if any) — plus a `meta` payload (capped logs, dispatch count) for presentation. A run with `result.error` throws a `CodeRunFailedError extends HarnessError` (`code: 'CODE_RUN_FAILED'`, message = the error kind and text plus captured logs so the model can self-correct); the registry's existing catch turns it into a structured `isError` result. +3. **Surfaces the outcome — after reaching quiescence.** When `ctx.codeRuntime.run()` resolves, the bridge fires the run-scoped abort (cancelling any in-flight sub-dispatch and abandoning queued-unstarted ones), then **awaits the dispatch queue's drain before returning**, per the dispose-to-quiescence rule in [defensive patterns](../../../defensive-patterns.md): an aborted in-flight sub-call still settles and logs its `isError` `tool/code-dispatch` event *inside* the open turn, and nothing can append after `run_code` returns. A successful run then returns one text block — the captured console/stdout output followed by the rendered return value (if any) — plus a `meta` payload (capped logs, dispatch count) for presentation. A run with `result.error` throws a `CodeRunFailedError extends HarnessError` (`code: 'CODE_RUN_FAILED'`, message = the error kind and text plus captured logs so the model can self-correct); the registry's existing catch turns it into a structured `isError` result. **Sub-call `additionalContext` is suppressed, deliberately.** A `tools/post-execute` hook may attach `additionalContext` to a call; for loop-dispatched calls the loop buffers those and appends each as a `context/message` only after the step's `tool/result`s, preserving call/result adjacency. A sub-dispatch result's `additionalContext` has no such safe outlet from inside a running `run_code`: injecting immediately would land a `context/message` between the parent's `tool/call` and its `tool/result` (breaking the adjacency the buffering exists to protect), and `PostToolDecision.additionalContext` is singular where a program may produce many. The MVP therefore drops sub-call `additionalContext`, pinned by a test and stated in the hooks bridge's docs; the follow-up (a plural context channel or loop-level sub-dispatch buffering) is deferred until a real hook needs it through Code Mode. @@ -117,7 +117,7 @@ The four PRs land in order (each on the previous); per stacked-review practice, - Under `mode: 'code'`, the assembled tool list (and thus the logged `request/header`) is exactly `[run_code]`; under `'both'`, every native schema plus `run_code`; the `tools:sdk` section is present in both, absent under `'native'`, and its text is deterministic for a fixed tool set (byte-identical across consecutive assemblies). - The generated `.d.ts` covers every registered tool except `run_code`, non-identifier names via quoted keys, unsupported schema constructs as `unknown`, without codegen ever throwing. - A program calling two tools returns only its curated output; each sub-call appears as a `tool/code-dispatch` event ordered by log `seq`, flows through `tools/pre-execute`/`post-execute` (a deny reaches the program as a binding rejection), and never enters derived messages. -- `Promise.all` over three bindings produces non-overlapping `ctx.tools.execute()` intervals (probe-tool assertion); aborting mid-program stops the worker and dispatches nothing further. +- `Promise.all` over three bindings produces non-overlapping `ctx.tools.execute()` intervals (probe-tool assertion); aborting mid-program stops the worker and dispatches nothing further; a budget expiry during a slow sub-dispatch aborts that dispatch (the probe tool observes its signal fire), `run_code` returns only after the queue drains, and no `tool/code-dispatch` event lands after `run_code`'s own `tool/result` in the log. - Worker runtime: a hot `for(;;){}` run ends at the `computeMs` busy-time budget with `error.kind: 'timeout'` — including when the program fired an un-awaited binding call first (the pending-RPC decoy); a program idling on a slow binding does not consume `computeMs` and is bounded only by `maxWallMs`; OOM under `resourceLimits` yields `kind: 'worker-exit'` with the host process healthy; `process.env` inside a program is empty; non-erasable syntax yields `kind: 'exception'` without a worker spawn; disposal awaits worker exit. - Misconfiguration is loud before any model request: non-native mode with no `ctx.codeRuntime`, a runtime whose `language ≠ 'typescript'`, and `toolOrder` naming a non-contributed tool all reject the assembly with actionable messages. - The demo runs against the real API via `demo:code`; the snapshot goldens replay keylessly; all repo gates pass on every PR of the stack. From 56b05f70cf8f67ac502a2888b0c97aa556517777 Mon Sep 17 00:00:00 2001 From: Tianyi Cui <53024+tianyicui@users.noreply.github.com> Date: Wed, 8 Jul 2026 02:17:24 +0800 Subject: [PATCH 14/26] feat: add the code-execution capability seam (ctx.codeRuntime) New group packages/code-runtime/ with the interface package @deepseek-ai/dsh-code-runtime, per the Code Mode RFC: abstract CodeRuntime service (run() resolves program failures as an error field, rejects only for seam misuse), the CodeRunRequest/CodeBindingNamespace/CodeRunResult/ CodeLogEntry/CodeRunFailure vocabulary, and readonly language/isolation backend descriptors. Registered in the tsconfig maps, packages/README, architecture service map, and the doc-graph service-role classification; catalogs regenerated. The RFC's one forward path token to the worker package becomes an npm-name mention until PR3 creates that directory (verify-package-paths is drift-scoped: the now-existing group made the token checkable). docs/architecture.md ceiling 1630 -> 1640: the doc gained a genuinely new capability-service row; the row itself is already minimal. --- docs/architecture.md | 1 + docs/capability-seams.md | 4 + docs/config-catalog.md | 1 + docs/cordis-catalog/services.md | 17 +++ docs/module-graph.md | 4 + .../proposed/feature/2026-06-15-code-mode.md | 2 +- packages/README.md | 1 + packages/code-runtime/README.md | 9 ++ packages/code-runtime/code-runtime/README.md | 19 ++++ .../code-runtime/code-runtime/package.json | 30 +++++ .../code-runtime/code-runtime/src/index.ts | 93 ++++++++++++++++ .../code-runtime/code-runtime/src/types.ts | 105 ++++++++++++++++++ .../code-runtime/tests/service.spec.ts | 87 +++++++++++++++ .../code-runtime/code-runtime/tsconfig.json | 18 +++ pnpm-lock.yaml | 6 + scripts/doc-budgets.manifest.json | 2 +- scripts/gen-doc-graphs.ts | 9 ++ tsconfig.base.json | 1 + tsconfig.build.json | 1 + tsconfig.json | 1 + 20 files changed, 409 insertions(+), 2 deletions(-) create mode 100644 packages/code-runtime/README.md create mode 100644 packages/code-runtime/code-runtime/README.md create mode 100644 packages/code-runtime/code-runtime/package.json create mode 100644 packages/code-runtime/code-runtime/src/index.ts create mode 100644 packages/code-runtime/code-runtime/src/types.ts create mode 100644 packages/code-runtime/code-runtime/tests/service.spec.ts create mode 100644 packages/code-runtime/code-runtime/tsconfig.json diff --git a/docs/architecture.md b/docs/architecture.md index 371b5df579..02dac4a2dd 100644 --- a/docs/architecture.md +++ b/docs/architecture.md @@ -26,6 +26,7 @@ The default distribution is a composition, not a hierarchy. `packages/core/` is |---|---|---| | `ctx.llm` | [`llm/`](../packages/llm/README.md) | adapter registry and streaming model calls | | `ctx.bash` | [`bash/`](../packages/bash/README.md) | foreground/background command execution | +| `ctx.codeRuntime` | [`code-runtime/`](../packages/code-runtime/README.md) | model-written program execution | | `ctx.fs` | [`fs/`](../packages/fs/README.md) | filesystem provider primitives and policy events | | `ctx.web` | [`web/`](../packages/web/README.md) | search/fetch provider registries | | `ctx.compact` | [`compact/`](../packages/compact/README.md) | session-surface compaction | diff --git a/docs/capability-seams.md b/docs/capability-seams.md index 414ee898d1..339954c00f 100644 --- a/docs/capability-seams.md +++ b/docs/capability-seams.md @@ -42,6 +42,8 @@ flowchart LR pkg_bash_local["bash-local"] pkg_hooks_claude["hooks-claude"] pkg_hooks_codex["hooks-codex"] + pkg_code_runtime["code-runtime"] + svc_codeRuntime["ctx.codeRuntime
Code-execution seam"] pkg_fs["fs"] svc_fs["ctx.fs
Filesystem provider seam"] pkg_fs_local["fs-local"] @@ -64,6 +66,7 @@ flowchart LR pkg_agent_loop --> svc_agentLoop pkg_bash --> svc_bash pkg_bash_local --> svc_bash + pkg_code_runtime --> svc_codeRuntime pkg_compact --> svc_compact pkg_compact_basic --> svc_compact pkg_fs --> svc_fs @@ -134,6 +137,7 @@ flowchart LR | `ctx.agents` | `core` | [`agent`](../packages/core/agent) | - | [`agent-loop`](../packages/core/agent-loop), [`acp`](../packages/ui/acp), [`subagent-inprocess`](../packages/subagent/subagent-inprocess), [`stdio-agent`](../packages/ui/stdio-agent), [`invariants`](../packages/support/invariants) | - | Owns live Agent handles and the create/resume factory seam. | | `ctx.agentLoop` | `bundle` | [`agent-loop`](../packages/core/agent-loop) | - | [`agent-core`](../packages/core/agent-core) | - | The one concrete loop plugin; extension packages depend on dsh-agent events and services, not on this package. | | `ctx.bash` | `seam` | [`bash`](../packages/bash/bash) | [`bash-local`](../packages/bash/bash-local) | [`tool-bash`](../packages/bash/tool-bash), [`hooks-claude`](../packages/hooks/hooks-claude), [`hooks-codex`](../packages/hooks/hooks-codex) | - | The model-facing bash tools and hook bridges consume this seam; sandboxed or remote executors can replace bash-local. | +| `ctx.codeRuntime` | `seam` | [`code-runtime`](../packages/code-runtime/code-runtime) | - | - | - | Runs one model-written program against host-provided async bindings; backends differ by substrate and language (the Code Mode RFC specifies the worker-thread backend and the tool-registry consumer). | | `ctx.fs` | `seam` | [`fs`](../packages/fs/fs) | [`fs-local`](../packages/fs/fs-local) | [`tool-fs`](../packages/fs/tool-fs) | [`fs-policy`](../packages/fs/fs-policy) | tool-fs executes read/write/edit through ctx.fs; fs-policy contributes observed-state checks through the fs/* event gate. | | `ctx.compact` | `seam` | [`compact`](../packages/compact/compact) | [`compact-basic`](../packages/compact/compact-basic) | [`compact-basic`](../packages/compact/compact-basic) | - | The basic backend currently consumes the pre-step event directly; a model-facing compact tool remains deferred. | | `ctx.subagents` | `seam` | [`subagent`](../packages/subagent/subagent) | [`subagent-spawn`](../packages/subagent/subagent-spawn), [`subagent-fork`](../packages/subagent/subagent-fork), [`subagent-acp`](../packages/subagent/subagent-acp), [`subagent-mock`](../packages/support/subagent-mock) | [`tool-subagent`](../packages/subagent/tool-subagent) | - | Providers implement transports; tool-subagent exposes one configured provider as a model-facing tool name. | diff --git a/docs/config-catalog.md b/docs/config-catalog.md index 5e8c5258e8..9105a52f6b 100644 --- a/docs/config-catalog.md +++ b/docs/config-catalog.md @@ -845,6 +845,7 @@ These load from a `cordis.yml` entry with no `config:` block; they declare no co Abstract service classes — a deployment loads a concrete implementation package instead ([capability seams](rfc/implemented/architecture/2026-06-13-capability-seams.md)). - `@deepseek-ai/dsh-bash` — abstract `BashExecutor` ([`packages/bash/bash/src/index.ts`](../packages/bash/bash/src/index.ts)) +- `@deepseek-ai/dsh-code-runtime` — abstract `CodeRuntime` ([`packages/code-runtime/code-runtime/src/index.ts`](../packages/code-runtime/code-runtime/src/index.ts)) - `@deepseek-ai/dsh-compact` — abstract `CompactService` ([`packages/compact/compact/src/index.ts`](../packages/compact/compact/src/index.ts)) - `@deepseek-ai/dsh-fs` — abstract `FileSystem` ([`packages/fs/fs/src/index.ts`](../packages/fs/fs/src/index.ts)) - `@deepseek-ai/dsh-session-persistence` — abstract `SessionPersistence` ([`packages/session-persistence/session-persistence/src/index.ts`](../packages/session-persistence/session-persistence/src/index.ts)) diff --git a/docs/cordis-catalog/services.md b/docs/cordis-catalog/services.md index f7f2734a9f..c941192381 100644 --- a/docs/cordis-catalog/services.md +++ b/docs/cordis-catalog/services.md @@ -67,6 +67,23 @@ Types: [BashExecRequest](../core-data-structures/bash.md) · [BashExecSpec](../c Source: [`packages/bash/bash/src/index.ts:59`](../../packages/bash/bash/src/index.ts) +## `ctx.codeRuntime` — `CodeRuntime` (abstract seam) + +Abstract code-execution service. Subclass, implement run and the two descriptors, and load the subclass as a plugin — it registers as `ctx.codeRuntime` (one implementation per context; loading a second throws, cordis' standard duplicate-service behavior). + +Semantics every implementation must honor: + +- run resolves with an error FIELD for every program outcome — parse/transform failures, thrown exceptions, budget expiry, abort, substrate death (CodeRunFailure's taxonomy). It REJECTS only for caller misuse of the seam itself (e.g. a run submitted after disposal). +- Binding calls bridge to the caller's CodeBindingFunctions verbatim; arguments and resolutions must be structured-cloneable, and the runtime treats the program as a hostile peer (arbitrary binding names are own properties, malformed traffic is rejected or ignored, never crashes the host). +- Runs are isolated from each other: no state survives from one run to the next through the runtime. +- Disposal reaches quiescence: in-flight runs are terminated AND awaited before the service's own teardown completes (no orphan substrate survives `fiber.dispose()`). + +```ts cordis-catalog +abstract run(request: CodeRunRequest): Promise +``` + +Source: [`packages/code-runtime/code-runtime/src/index.ts:59`](../../packages/code-runtime/code-runtime/src/index.ts) + ## `ctx.compact` — `CompactService` (abstract seam) Abstract compaction service. Subclass implement the two abstract methods, and load the subclass as a plugin — it registers as `ctx.compact` (one implementation per context; loading a second throws, which is cordis' standard duplicate-service behavior). diff --git a/docs/module-graph.md b/docs/module-graph.md index 81b272839d..29d454c68d 100644 --- a/docs/module-graph.md +++ b/docs/module-graph.md @@ -82,6 +82,9 @@ flowchart TD subgraph group_mcp["packages/mcp"] pkg_mcp_client["mcp-client"] end + subgraph group_code_runtime["packages/code-runtime"] + pkg_code_runtime["code-runtime"] + end pkg_llm --> pkg_brand pkg_bash --> pkg_brand pkg_llm_deepseek --> pkg_llm @@ -215,6 +218,7 @@ flowchart TD | [`brand`](../packages/util/brand) | `util` | — | | [`acp-snapshot`](../packages/support/acp-snapshot) | `support` | — | | [`app-boot`](../packages/ui/app-boot) | `ui` | — | +| [`code-runtime`](../packages/code-runtime/code-runtime) | `code-runtime` | — | | [`llm`](../packages/llm/llm) | `llm` | [`brand`](../packages/util/brand) | | [`bash`](../packages/bash/bash) | `bash` | [`brand`](../packages/util/brand) | | [`llm-deepseek`](../packages/llm/llm-deepseek) | `llm` | [`llm`](../packages/llm/llm) | diff --git a/docs/rfc/proposed/feature/2026-06-15-code-mode.md b/docs/rfc/proposed/feature/2026-06-15-code-mode.md index 30f4d71319..a5f1563126 100644 --- a/docs/rfc/proposed/feature/2026-06-15-code-mode.md +++ b/docs/rfc/proposed/feature/2026-06-15-code-mode.md @@ -65,7 +65,7 @@ Per explicit-over-implicit at seams, the request spells out everything the runti ### The worker-thread runtime -`packages/code-runtime/code-runtime-worker/` — `@deepseek-ai/dsh-code-runtime-worker`. Per `run()`: +`@deepseek-ai/dsh-code-runtime-worker`, the second package of the `packages/code-runtime/` group. Per `run()`: 1. **Type-strip host-side** with Node's built-in `stripTypeScriptTypes` (`node:module`; present across the repo's whole engines range, `^22.19.0 || >=24.0.0`, and position-preserving, so runtime error line numbers match the model's source). Strip-only mode rejects non-erasable syntax (`enum`, namespaces) — that rejection returns as `error.kind: 'exception'` with Node's message, the SDK instructions say "erasable TypeScript only", and the model self-corrects like any other program error. A syntax-level failure never spawns a worker. 2. **Spawn one fresh `Worker` per run** from the package's own bootstrap module: `env: {}` (truly empty — stronger than the scrubbed-env rule for spawned commands), `resourceLimits` from config, `stdout`/`stderr` captured into `logs` rather than inherited. No pooling and no cross-run state: a program's world dies with its worker, which keeps runs reconstructable from the log alone and makes state bleed unrepresentable. diff --git a/packages/README.md b/packages/README.md index f07e0d5a36..da75f740e8 100644 --- a/packages/README.md +++ b/packages/README.md @@ -11,6 +11,7 @@ Packages are grouped by modular role at `packages///`. The group dir | [`core/`](core/README.md) | Product API spine: session, system-prompt, tools, agent, and the concrete loop | Product — stable surface | | [`llm/`](llm/README.md) | LLM capability family: the abstract service + provider adapters | Product — stable surface | | [`bash/`](bash/README.md) | Bash capability family: the executor seam, a local impl, and the model-facing tool | Product — stable surface | +| [`code-runtime/`](code-runtime/README.md) | Code-execution capability family: the abstract runtime seam for model-written programs | Product — stable surface | | [`fs/`](fs/README.md) | Filesystem capability family: the abstract seam, a local impl, and the model-facing file tools | Product — stable surface | | [`compact/`](compact/README.md) | Compaction capability family: the abstract seam + a basic backend (tool deferred) | Product — stable surface | | [`subagent/`](subagent/README.md) | Subagent capability family: the provider-registry seam and the model-facing delegation tool | Product — stable surface | diff --git a/packages/code-runtime/README.md b/packages/code-runtime/README.md new file mode 100644 index 0000000000..578f3179c1 --- /dev/null +++ b/packages/code-runtime/README.md @@ -0,0 +1,9 @@ +# code-runtime/ — code-execution capability family + +The code-execution capability seam (see [capability seams](../../docs/rfc/implemented/architecture/2026-06-13-capability-seams.md)): an abstract runtime interface for executing one model-written program against host-provided async bindings, capturing what it printed and returned. The consumer is the tool registry's Code Mode, and the first implementation (a Node worker-thread backend) is specified alongside it in the [Code Mode RFC](../../docs/rfc/proposed/feature/2026-06-15-code-mode.md). **Product** packages. + +| Package | Role | ctx key | +|---|---|---| +| `code-runtime/` | Abstract code-execution seam (interface + vocabulary) | `ctx.codeRuntime` | + +The interface lives at `code-runtime/code-runtime/`. Backends differ by execution substrate (worker thread, process, container) and by source language — both readonly descriptors on the service — and register `ctx.codeRuntime` without touching the interface or its consumer; that split is what makes a hardened backend a drop-in later. diff --git a/packages/code-runtime/code-runtime/README.md b/packages/code-runtime/code-runtime/README.md new file mode 100644 index 0000000000..2d7b12add1 --- /dev/null +++ b/packages/code-runtime/code-runtime/README.md @@ -0,0 +1,19 @@ +# @deepseek-ai/dsh-code-runtime + +The **code-execution seam**: an abstract `CodeRuntime` service (`ctx.codeRuntime`) defining WHAT a code runtime does — run one model-written program against a set of host-provided async bindings and report `{ value, logs, error? }` — without saying HOW. + +This package is the interface third of the capability (the bash trio is the template — see [capability seams](../../../docs/rfc/implemented/architecture/2026-06-13-capability-seams.md)): implementations subclass `CodeRuntime` and register the service; the consumer is the tool registry's Code Mode, which generates the model-facing SDK and bridges tool dispatch — both specified in the [Code Mode RFC](../../../docs/rfc/proposed/feature/2026-06-15-code-mode.md), whose first implementation is a Node worker-thread backend. The runtime knows nothing about tools or sessions: it is handed named async functions and a program string, and everything tool-shaped stays with the consumer. + +## Service API (`ctx.codeRuntime`) + +| Member | Semantics | +|---|---| +| `run(request)` | Execute one program against the request's bindings. **Resolves with an error FIELD for every program outcome** — parse/transform failure, thrown exception, budget expiry, abort, substrate death (`CodeRunFailure`'s orthogonal `kind` taxonomy); it rejects only for caller misuse of the seam itself (e.g. a run submitted after disposal). The program runs as the body of an async function: top-level `await`/`return` work, and the completion value becomes `result.value` when it survives the serialization boundary. | +| `language` | Readonly descriptor: the source language `run` expects (`'typescript'` is the well-known value). Informational, not gating — a consumer that generates language-specific presentation switches on it and fails loud on a language it cannot present. | +| `isolation` | Readonly descriptor: the execution substrate (`'worker-thread'`, `'process'`, `'container'`). A label for deployments and diagnostics, **not a security claim**. | + +Semantics every implementation must honor (contract details in the class JSDoc): binding calls bridge to the caller's functions verbatim with structured-cloneable arguments/resolutions; the program is treated as a hostile peer (arbitrary binding names are own properties, malformed traffic never crashes the host); no state survives between runs; disposal terminates in-flight runs AND awaits their exit before completing. + +## Vocabulary + +`CodeRunRequest` (`program`, `bindings`, `signal?`) carries everything the runtime acts on — defaulting (time budgets, output caps) is the implementation's validated config, never a hidden `??` inside `run()`. `bindings` is a list of `CodeBindingNamespace`s (`global` + `functions`), each exposed to the program as one global object of async callables. `CodeRunResult` reports the completion `value?`, the ordered `logs` (`CodeLogEntry`: `console`/`stdout`/`stderr` source, console `level`, capped text), and the `error?` (`CodeRunFailure`: `kind` + model-feedable `message`). See `src/types.ts` for the full contracts. diff --git a/packages/code-runtime/code-runtime/package.json b/packages/code-runtime/code-runtime/package.json new file mode 100644 index 0000000000..0fe24bb15c --- /dev/null +++ b/packages/code-runtime/code-runtime/package.json @@ -0,0 +1,30 @@ +{ + "name": "@deepseek-ai/dsh-code-runtime", + "description": "Abstract code-execution seam (ctx.codeRuntime) for the DeepSeek Harness", + "version": "0.0.1", + "private": true, + "type": "module", + "main": "lib/index.js", + "types": "lib/types/index.d.ts", + "exports": { + ".": { + "types": "./lib/types/index.d.ts", + "default": "./lib/index.js" + }, + "./src/*": "./src/*", + "./package.json": "./package.json" + }, + "files": [ + "lib/index.js", + "lib/types/**/*.d.ts", + "lib/types/**/*.d.ts.map", + "src" + ], + "license": "BSD-3-Clause", + "peerDependencies": { + "cordis": "^4.0.0-rc.6" + }, + "devDependencies": { + "cordis": "^4.0.0-rc.6" + } +} diff --git a/packages/code-runtime/code-runtime/src/index.ts b/packages/code-runtime/code-runtime/src/index.ts new file mode 100644 index 0000000000..af967da61d --- /dev/null +++ b/packages/code-runtime/code-runtime/src/index.ts @@ -0,0 +1,93 @@ +/** + * The code-execution seam (`ctx.codeRuntime`): an abstract service defining + * WHAT a code runtime does — run one model-written program against a set of + * host-provided async bindings and report `{ value, logs, error? }` — without + * saying HOW. Implementations subclass {@link CodeRuntime} and register + * themselves as the `codeRuntime` service; backends may differ by execution + * substrate (worker thread, separate process, container) and by source + * language, both declared as readonly descriptors. The design and its + * consumer (the tool registry's Code Mode) are specified in the Code Mode RFC + * (docs/rfc/proposed/feature/2026-06-15-code-mode.md). + * + * The split mirrors the bash seam (`BashExecutor`): the runtime knows nothing + * about tools or sessions — it is handed named async functions and a program, + * and everything tool-shaped stays with the consumer. + * + * @module @deepseek-ai/dsh-code-runtime + */ + +import { Context, Service } from 'cordis' +import type { CodeRunRequest, CodeRunResult } from './types.ts' + +export type { + CodeBindingFunction, + CodeBindingNamespace, + CodeLogEntry, + CodeRunFailure, + CodeRunRequest, + CodeRunResult, +} from './types.ts' + +declare module 'cordis' { + interface Context { + codeRuntime: CodeRuntime + } +} + +/** + * Abstract code-execution service. Subclass, implement {@link run} and the + * two descriptors, and load the subclass as a plugin — it registers as + * `ctx.codeRuntime` (one implementation per context; loading a second throws, + * cordis' standard duplicate-service behavior). + * + * Semantics every implementation must honor: + * - {@link run} resolves with an error FIELD for every program outcome — + * parse/transform failures, thrown exceptions, budget expiry, abort, + * substrate death ({@link CodeRunFailure}'s taxonomy). It REJECTS only for + * caller misuse of the seam itself (e.g. a run submitted after disposal). + * - Binding calls bridge to the caller's {@link CodeBindingFunction}s + * verbatim; arguments and resolutions must be structured-cloneable, and the + * runtime treats the program as a hostile peer (arbitrary binding names are + * own properties, malformed traffic is rejected or ignored, never crashes + * the host). + * - Runs are isolated from each other: no state survives from one run to the + * next through the runtime. + * - Disposal reaches quiescence: in-flight runs are terminated AND awaited + * before the service's own teardown completes (no orphan substrate survives + * `fiber.dispose()`). + */ +export abstract class CodeRuntime extends Service { + /** + * The source language {@link run} expects `program` to be written in, as a + * lowercase identifier. Informational, not gating — a consumer that + * generates language-specific presentation (typed SDK stubs, usage + * instructions) switches on it and fails loud on a language it cannot + * present. Well-known value: `'typescript'`. + */ + abstract readonly language: string + + /** + * The execution substrate, as a lowercase identifier. Informational, not + * gating — a descriptor so deployments and diagnostics can tell backends + * apart, not a security claim. Well-known values: `'worker-thread'`, + * `'process'`, `'container'`. + */ + abstract readonly isolation: string + + constructor(ctx: Context) { + super(ctx, 'codeRuntime') + } + + /** + * Execute one program against the request's bindings and capture what it + * emitted. See the class doc for the resolution contract (error is a result + * field; rejection means seam misuse only). + * @param request - the program, its bindings, and the abort signal; the + * request carries everything the runtime acts on, with no hidden defaults. + * @returns the run's outcome: completion value (when transferable), the + * ordered log capture, and the failure (if any). + */ + abstract run(request: CodeRunRequest): Promise +} + +export default CodeRuntime diff --git a/packages/code-runtime/code-runtime/src/types.ts b/packages/code-runtime/code-runtime/src/types.ts new file mode 100644 index 0000000000..8278f33a39 --- /dev/null +++ b/packages/code-runtime/code-runtime/src/types.ts @@ -0,0 +1,105 @@ +/** + * Vocabulary types for the code-execution seam: what a caller hands a + * {@link ../index.ts | CodeRuntime} and what it gets back. Pure types — no + * runtime code lives here. + * + * @module @deepseek-ai/dsh-code-runtime/src/types + */ + +/** + * One host-side function exposed to the program as an async callable. The + * runtime bridges calls to it (possibly across a serialization boundary), so + * `args` and the resolution value MUST be structured-cloneable; a runtime + * rejects a non-cloneable value with a descriptive error rather than + * corrupting the run. A rejection of this function surfaces inside the + * program as a rejection of the corresponding call. + */ +export type CodeBindingFunction = (args: unknown) => Promise + +/** + * A named group of {@link CodeBindingFunction}s the runtime exposes to the + * program as one global object (e.g. `tools`). Function names are arbitrary + * strings — a runtime must treat names like `__proto__` or `constructor` as + * ordinary own properties (null-prototype construction), never as prototype + * collisions. + */ +export interface CodeBindingNamespace { + /** The global identifier the program sees (must be a valid JS identifier). */ + global: string + /** The callable members, keyed by the exact name the program calls. */ + functions: Record +} + +/** + * One run: the program source plus everything the runtime acts on. Per the + * explicit-over-implicit convention, defaulting (time budgets, output caps) + * is the implementation's validated config — a request carries no optional + * tuning knobs for a hidden `??` to fill in. + */ +export interface CodeRunRequest { + /** + * The program source, in the runtime's {@link ../index.ts | language}. It + * runs as the body of an async function: top-level `await` and `return` + * are available, and the completion value becomes + * {@link CodeRunResult.value}. + */ + program: string + /** Host functions exposed to the program, one global object per namespace. */ + bindings: CodeBindingNamespace[] + /** + * Abort the run: the runtime stops the program (hard, even mid-loop) and + * resolves with a {@link CodeRunFailure} of kind `'abort'`. In-flight + * binding calls are the CALLER's to settle — the runtime only stops asking. + */ + signal?: AbortSignal +} + +/** + * One captured output entry, in emission order. `source` says which channel + * produced it: the program's `console` (shimmed by the runtime), or a stray + * write to the underlying stdout/stderr streams. + */ +export interface CodeLogEntry { + /** Which channel produced the text. */ + source: 'console' | 'stdout' | 'stderr' + /** The console method used; present only when `source` is `'console'`. */ + level?: 'log' | 'info' | 'warn' | 'error' | 'debug' + /** The captured text (possibly truncated by the implementation's caps, marked in-band). */ + text: string +} + +/** + * Why a run failed. The kinds are orthogonal outcomes reported independently + * (per docs/defensive-patterns.md): a budget expiry is not an exception, an + * abort is not a timeout, and a substrate death is neither. + * + * - `'exception'` — the program threw or failed to parse/transform. + * - `'timeout'` — an implementation-owned budget expired; the message says which. + * - `'abort'` — {@link CodeRunRequest.signal} fired. + * - `'worker-exit'` — the execution substrate died without settling (e.g. OOM). + */ +export interface CodeRunFailure { + /** The failure class (see the interface doc for each kind's meaning). */ + kind: 'exception' | 'timeout' | 'abort' | 'worker-exit' + /** Human-readable detail, suitable for feeding back to a model to self-correct. */ + message: string +} + +/** + * The outcome of one run. An error is a FIELD on a resolved result, never a + * rejection of `run()` — reporting a failed program is the caller's job, not + * an exception path. + */ +export interface CodeRunResult { + /** + * The program's completion value (its top-level `return`), when it ran to + * completion and the value survived the runtime's serialization boundary; + * a non-transferable value is replaced by a string rendering, and a failed + * or value-less run leaves this absent. + */ + value?: unknown + /** Everything the program emitted, in order (capped by the implementation). */ + logs: CodeLogEntry[] + /** Present iff the run failed; see {@link CodeRunFailure} for the taxonomy. */ + error?: CodeRunFailure +} diff --git a/packages/code-runtime/code-runtime/tests/service.spec.ts b/packages/code-runtime/code-runtime/tests/service.spec.ts new file mode 100644 index 0000000000..4ff6d8f313 --- /dev/null +++ b/packages/code-runtime/code-runtime/tests/service.spec.ts @@ -0,0 +1,87 @@ +import { describe, expect, it } from 'vitest' +import { Context } from 'cordis' +import { CodeRuntime } from '@deepseek-ai/dsh-code-runtime' +import type { CodeRunRequest, CodeRunResult } from '@deepseek-ai/dsh-code-runtime' + +/** + * Minimal concrete runtime: records requests, "executes" by invoking every + * binding once in declaration order, and lets tests script the outcome. The + * seam package ships no implementation, so the contract is exercised through + * the smallest subclass that honors it. + */ +class StubRuntime extends CodeRuntime { + readonly language = 'typescript' + readonly isolation = 'in-process-stub' + requests: CodeRunRequest[] = [] + nextResult: CodeRunResult = { logs: [] } + + async run(request: CodeRunRequest): Promise { + this.requests.push(request) + if (request.signal?.aborted) { + return { logs: [], error: { kind: 'abort', message: String(request.signal.reason) } } + } + for (const namespace of request.bindings) { + for (const fn of Object.values(namespace.functions)) { + await fn({ from: 'stub' }) + } + } + return this.nextResult + } +} + +async function setup() { + const ctx = new Context() + await ctx.plugin(StubRuntime) + const runtime = ctx.codeRuntime as StubRuntime + return { ctx, runtime } +} + +describe('CodeRuntime service seam', () => { + it('registers as ctx.codeRuntime and serves the abstract API', async () => { + const { runtime } = await setup() + expect(runtime.language).toBe('typescript') + expect(runtime.isolation).toBe('in-process-stub') + + const calls: unknown[] = [] + const result = await runtime.run({ + program: 'return 1', + bindings: [{ global: 'tools', functions: { probe: async args => void calls.push(args) } }], + }) + expect(result).toEqual({ logs: [] }) + expect(calls).toEqual([{ from: 'stub' }]) + expect(runtime.requests).toHaveLength(1) + }) + + it('reports a failed run as an error field on a resolved result, never a rejection', async () => { + const { runtime } = await setup() + runtime.nextResult = { + logs: [{ source: 'console', level: 'error', text: 'boom' }], + error: { kind: 'exception', message: 'boom' }, + } + const result = await runtime.run({ program: 'throw new Error("boom")', bindings: [] }) + expect(result.error).toEqual({ kind: 'exception', message: 'boom' }) + expect(result.value).toBeUndefined() + }) + + it('reports a pre-aborted signal as an abort failure', async () => { + const { runtime } = await setup() + const controller = new AbortController() + controller.abort('cancelled') + const result = await runtime.run({ program: 'return 1', bindings: [], signal: controller.signal }) + expect(result.error).toEqual({ kind: 'abort', message: 'cancelled' }) + }) + + it('is removed from the context when the providing fiber disposes (HMR safety)', async () => { + const ctx = new Context() + const fiber = await ctx.plugin(StubRuntime) + expect(ctx.get('codeRuntime')).toBeInstanceOf(StubRuntime) + + await fiber.dispose() + expect(ctx.get('codeRuntime')).toBeUndefined() + }) + + it('rejects a second implementation in the same context (duplicate service)', async () => { + const { ctx } = await setup() + await expect(ctx.plugin(StubRuntime)).rejects.toThrow(/registered/) + }) +}) diff --git a/packages/code-runtime/code-runtime/tsconfig.json b/packages/code-runtime/code-runtime/tsconfig.json new file mode 100644 index 0000000000..754725418e --- /dev/null +++ b/packages/code-runtime/code-runtime/tsconfig.json @@ -0,0 +1,18 @@ +{ + "extends": "../../../tsconfig.base.json", + "compilerOptions": { + "rootDir": "src", + "outDir": "lib/types" + }, + "include": [ + "src" + ], + "references": [ + { + "path": "../../../vendor/cosmokit" + }, + { + "path": "../../../vendor/cordis" + } + ] +} diff --git a/pnpm-lock.yaml b/pnpm-lock.yaml index d12ec0d9c5..cd6918087d 100644 --- a/pnpm-lock.yaml +++ b/pnpm-lock.yaml @@ -127,6 +127,12 @@ importers: specifier: ^4.0.0-rc.6 version: 4.0.0-rc.6(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.4) + packages/code-runtime/code-runtime: + devDependencies: + cordis: + specifier: ^4.0.0-rc.6 + version: 4.0.0-rc.6(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.4) + packages/compact/compact: devDependencies: '@deepseek-ai/dsh-llm': diff --git a/scripts/doc-budgets.manifest.json b/scripts/doc-budgets.manifest.json index b23811e3d0..fc2b9d12c2 100644 --- a/scripts/doc-budgets.manifest.json +++ b/scripts/doc-budgets.manifest.json @@ -1,7 +1,7 @@ { "AGENTS.md": 1691, "docs/AGENTS.md": 1315, - "docs/architecture.md": 1630, + "docs/architecture.md": 1640, "docs/cordis-primer.md": 550, "docs/defensive-patterns.md": 550, "docs/testing.md": 800, diff --git a/scripts/gen-doc-graphs.ts b/scripts/gen-doc-graphs.ts index 66cb102b00..6440a672be 100644 --- a/scripts/gen-doc-graphs.ts +++ b/scripts/gen-doc-graphs.ts @@ -149,6 +149,15 @@ const SERVICE_ROLES: ServiceRole[] = [ consumers: ['tool-bash', 'hooks-claude', 'hooks-codex'], note: 'The model-facing bash tools and hook bridges consume this seam; sandboxed or remote executors can replace bash-local.', }, + { + key: 'codeRuntime', + pkg: 'code-runtime', + title: 'Code-execution seam', + mode: 'seam', + implementations: [], + consumers: [], + note: 'Runs one model-written program against host-provided async bindings; backends differ by substrate and language (the Code Mode RFC specifies the worker-thread backend and the tool-registry consumer).', + }, { key: 'fs', pkg: 'fs', diff --git a/tsconfig.base.json b/tsconfig.base.json index 7cbb3acab2..a38b4e1836 100644 --- a/tsconfig.base.json +++ b/tsconfig.base.json @@ -43,6 +43,7 @@ "./packages/core/*/src", "./packages/llm/*/src", "./packages/bash/*/src", + "./packages/code-runtime/*/src", "./packages/fs/*/src", "./packages/compact/*/src", "./packages/subagent/*/src", diff --git a/tsconfig.build.json b/tsconfig.build.json index b44368db21..ee62bca604 100644 --- a/tsconfig.build.json +++ b/tsconfig.build.json @@ -22,6 +22,7 @@ { "path": "./packages/core/agent-loop" }, { "path": "./packages/core/agent-core" }, { "path": "./packages/bash/bash" }, + { "path": "./packages/code-runtime/code-runtime" }, { "path": "./packages/compact/compact" }, { "path": "./packages/compact/compact-basic" }, { "path": "./packages/llm/llm-deepseek" }, diff --git a/tsconfig.json b/tsconfig.json index b257b0aa89..8a85a4cd59 100644 --- a/tsconfig.json +++ b/tsconfig.json @@ -33,6 +33,7 @@ { "path": "./packages/core/agent-loop" }, { "path": "./packages/core/agent-core" }, { "path": "./packages/bash/bash" }, + { "path": "./packages/code-runtime/code-runtime" }, { "path": "./packages/llm/llm-deepseek" }, { "path": "./packages/llm/llm-pi-ai" }, { "path": "./packages/bash/bash-local" }, From 0ba6de00010b2e67ab945d9819a4f1dba1747886 Mon Sep 17 00:00:00 2001 From: Tianyi Cui <53024+tianyicui@users.noreply.github.com> Date: Wed, 8 Jul 2026 02:38:47 +0800 Subject: [PATCH 15/26] docs: catalog the code-runtime seam vocabulary (Codex review finding) Adds the missing core-data-structures coverage the catalog policy requires for non-spine seam vocabulary: the code-runtime.md sub-page with drift-checked type-equiv blocks for all six seam types, the core.md sub-page row, the type-equiv manifest entries, and LINK_MAP entries so the generated service signature links CodeRunRequest/CodeRunResult; cordis/config catalogs regenerated. --- docs/cordis-catalog/services.md | 2 + docs/core-data-structures/code-runtime.md | 94 +++++++++++++++++++++++ docs/core-data-structures/core.md | 1 + scripts/gen-cordis-catalog.ts | 2 + scripts/type-equiv.manifest.json | 7 ++ 5 files changed, 106 insertions(+) create mode 100644 docs/core-data-structures/code-runtime.md diff --git a/docs/cordis-catalog/services.md b/docs/cordis-catalog/services.md index c941192381..26126c4481 100644 --- a/docs/cordis-catalog/services.md +++ b/docs/cordis-catalog/services.md @@ -82,6 +82,8 @@ Semantics every implementation must honor: abstract run(request: CodeRunRequest): Promise ``` +Types: [CodeRunRequest](../core-data-structures/code-runtime.md) · [CodeRunResult](../core-data-structures/code-runtime.md) + Source: [`packages/code-runtime/code-runtime/src/index.ts:59`](../../packages/code-runtime/code-runtime/src/index.ts) ## `ctx.compact` — `CompactService` (abstract seam) diff --git a/docs/core-data-structures/code-runtime.md b/docs/core-data-structures/code-runtime.md new file mode 100644 index 0000000000..1f87e8e8a4 --- /dev/null +++ b/docs/core-data-structures/code-runtime.md @@ -0,0 +1,94 @@ +# Code Runtime + +The code-execution seam — a [capability seam](../rfc/implemented/architecture/2026-06-13-capability-seams.md) whose interface ([dsh-code-runtime](../../packages/code-runtime/code-runtime), `ctx.codeRuntime`) runs one model-written program against host-provided async bindings and reports what it printed and returned. Code execution is **one optional capability**, not part of the agent-loop spine — so its vocabulary lives here, not in [core.md](core.md). Backends differ by execution substrate and source language, both readonly descriptors on the service; the worker-thread backend and the tool-registry consumer (Code Mode) are specified in the [Code Mode RFC](../rfc/proposed/feature/2026-06-15-code-mode.md). + +Source: [`packages/code-runtime/code-runtime/src/types.ts`](../../packages/code-runtime/code-runtime/src/types.ts) + +## The run: request in, result out + +A `CodeRunRequest` carries **everything the runtime acts on** — per the "explicit > implicit at package seams" rule, defaulting (time budgets, output caps) is the implementation's validated config, never a hidden `??` inside `run()`: + +```ts type-equiv +interface CodeRunRequest { + /** + * The program source, in the runtime's {@link ../index.ts | language}. It + * runs as the body of an async function: top-level `await` and `return` + * are available, and the completion value becomes + * {@link CodeRunResult.value}. + */ + program: string + /** Host functions exposed to the program, one global object per namespace. */ + bindings: CodeBindingNamespace[] + /** + * Abort the run: the runtime stops the program (hard, even mid-loop) and + * resolves with a {@link CodeRunFailure} of kind `'abort'`. In-flight + * binding calls are the CALLER's to settle — the runtime only stops asking. + */ + signal?: AbortSignal +} +``` + +The result reports an error as a **field**, never a rejection of `run()` — reporting a failed program is the caller's job, not an exception path (mirroring `BashExecutor.run`'s resolve-on-failure contract): + +```ts type-equiv +interface CodeRunResult { + /** + * The program's completion value (its top-level `return`), when it ran to + * completion and the value survived the runtime's serialization boundary; + * a non-transferable value is replaced by a string rendering, and a failed + * or value-less run leaves this absent. + */ + value?: unknown + /** Everything the program emitted, in order (capped by the implementation). */ + logs: CodeLogEntry[] + /** Present iff the run failed; see {@link CodeRunFailure} for the taxonomy. */ + error?: CodeRunFailure +} +``` + +## Bindings: host functions as program globals + +Each `CodeBindingNamespace` becomes one global object of async callables inside the program (the Code Mode consumer passes one: `tools`). Arguments and resolutions must be structured-cloneable — a runtime may bridge calls across a serialization boundary — and a runtime treats binding names as hostile input (`__proto__` is an ordinary own property, never a prototype collision): + +```ts type-equiv +interface CodeBindingNamespace { + /** The global identifier the program sees (must be a valid JS identifier). */ + global: string + /** The callable members, keyed by the exact name the program calls. */ + functions: Record +} +``` + +```ts type-equiv +type CodeBindingFunction = (args: unknown) => Promise +``` + +## Captured output and the failure taxonomy + +Logs arrive in emission order, attributed to their channel (the runtime's `console` shim, or stray writes to the underlying streams): + +```ts type-equiv +interface CodeLogEntry { + /** Which channel produced the text. */ + source: 'console' | 'stdout' | 'stderr' + /** The console method used; present only when `source` is `'console'`. */ + level?: 'log' | 'info' | 'warn' | 'error' | 'debug' + /** The captured text (possibly truncated by the implementation's caps, marked in-band). */ + text: string +} +``` + +Failure kinds are **orthogonal outcomes reported independently** (per [defensive-patterns](../defensive-patterns.md)): a budget expiry is not an exception, an abort is not a timeout, and a substrate death (e.g. OOM) is neither: + +```ts type-equiv +interface CodeRunFailure { + /** The failure class (see the interface doc for each kind's meaning). */ + kind: 'exception' | 'timeout' | 'abort' | 'worker-exit' + /** Human-readable detail, suitable for feeding back to a model to self-correct. */ + message: string +} +``` + +## The service + +`CodeRuntime` (`ctx.codeRuntime`, abstract — defined in [`packages/code-runtime/code-runtime/src/index.ts`](../../packages/code-runtime/code-runtime/src/index.ts)) is `run(request)` plus two readonly descriptors: `language` (what the program must be written in — `'typescript'` is the well-known value; a consumer generating language-specific presentation switches on it and fails loud on one it cannot present) and `isolation` (the execution substrate — `'worker-thread'`, `'process'`, `'container'`; a diagnostic label, **not a security claim**). Implementations must keep runs isolated from each other (no cross-run state) and dispose to quiescence: in-flight runs are terminated and awaited before teardown completes. diff --git a/docs/core-data-structures/core.md b/docs/core-data-structures/core.md index 7d1110d7a8..615c222d94 100644 --- a/docs/core-data-structures/core.md +++ b/docs/core-data-structures/core.md @@ -20,6 +20,7 @@ Everything else is documented on a **sub-page**, not here. The rule that draws t | [persistence.md](persistence.md) | the durability seam: `SessionPersistence`, JSONL + SQLite backends, `session/flush`, crash recovery, `SessionHeader` | | [tools.md](tools.md) | `ToolDefinition` full fields, the schema DSL, `ToolExecution`/`ToolResult`, tool-presentation UI types, the `tools/pre-execute`/`tools/post-execute` pipeline | | [bash.md](bash.md) | the bash executor seam: `BashExecRequest`/`Spec`, `BashRunResult`, background `BashTask`s | +| [code-runtime.md](code-runtime.md) | the code-execution seam: `CodeRunRequest`/`Result`, binding namespaces, captured logs, the `CodeRunFailure` taxonomy | | [filesystem.md](filesystem.md) | the filesystem seam: `FsTarget`, read/write/edit outcomes, observed-file state, `FsErrorCode` | | [compaction.md](compaction.md) | the compaction seam: the `compact/*` session events, `CompactionResult`, the `CompactService` interface | | [subagent.md](subagent.md) | the subagent seam: the named-provider registry, `SubagentStartRequest`/`Result`/`Run`, the start-time-vs-runtime capability split | diff --git a/scripts/gen-cordis-catalog.ts b/scripts/gen-cordis-catalog.ts index 451c4bec9e..6220006ede 100644 --- a/scripts/gen-cordis-catalog.ts +++ b/scripts/gen-cordis-catalog.ts @@ -97,6 +97,8 @@ export const LINK_MAP: Record = { BashRunResult: 'bash.md', BashTask: 'bash.md', BashTaskRead: 'bash.md', + CodeRunRequest: 'code-runtime.md', + CodeRunResult: 'code-runtime.md', FsEditOutcome: 'filesystem.md', FsEditRequest: 'filesystem.md', FsInfo: 'filesystem.md', diff --git a/scripts/type-equiv.manifest.json b/scripts/type-equiv.manifest.json index b66882d8c9..613280ee1b 100644 --- a/scripts/type-equiv.manifest.json +++ b/scripts/type-equiv.manifest.json @@ -55,6 +55,13 @@ { "doc": "docs/core-data-structures/bash.md", "symbol": "BashTask", "source": "packages/bash/bash/src/types.ts" }, { "doc": "docs/core-data-structures/bash.md", "symbol": "BashTaskRead", "source": "packages/bash/bash/src/types.ts" }, + { "doc": "docs/core-data-structures/code-runtime.md", "symbol": "CodeRunRequest", "source": "packages/code-runtime/code-runtime/src/types.ts" }, + { "doc": "docs/core-data-structures/code-runtime.md", "symbol": "CodeRunResult", "source": "packages/code-runtime/code-runtime/src/types.ts" }, + { "doc": "docs/core-data-structures/code-runtime.md", "symbol": "CodeBindingNamespace", "source": "packages/code-runtime/code-runtime/src/types.ts" }, + { "doc": "docs/core-data-structures/code-runtime.md", "symbol": "CodeBindingFunction", "source": "packages/code-runtime/code-runtime/src/types.ts" }, + { "doc": "docs/core-data-structures/code-runtime.md", "symbol": "CodeLogEntry", "source": "packages/code-runtime/code-runtime/src/types.ts" }, + { "doc": "docs/core-data-structures/code-runtime.md", "symbol": "CodeRunFailure", "source": "packages/code-runtime/code-runtime/src/types.ts" }, + { "doc": "docs/core-data-structures/filesystem.md", "symbol": "FsTarget", "source": "packages/fs/fs/src/types.ts" }, { "doc": "docs/core-data-structures/filesystem.md", "symbol": "FsTargetKey", "source": "packages/fs/fs/src/types.ts" }, { "doc": "docs/core-data-structures/filesystem.md", "symbol": "FsVersion", "source": "packages/fs/fs/src/types.ts" }, From 80ce8b8dd47e70fdc21e90ae140114e050e1b23e Mon Sep 17 00:00:00 2001 From: Tianyi Cui <53024+tianyicui@users.noreply.github.com> Date: Wed, 8 Jul 2026 10:42:14 +0800 Subject: [PATCH 16/26] docs: pin JSON normalization at the dispatch bridge (review finding) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The seam's structured-clone boundary admits values JSON does not (BigInt, Map, circulars), while tool/code-dispatch events must be JSON-appendable — left unhandled, a sub-call could execute and then fail at logging time. The bridge now JSON-normalizes binding arguments BEFORE dispatch (a value that does not survive rejects that one call), so the dispatched form and the logged form are the same JSON value by construction. --- docs/rfc/proposed/feature/2026-06-15-code-mode.md | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/docs/rfc/proposed/feature/2026-06-15-code-mode.md b/docs/rfc/proposed/feature/2026-06-15-code-mode.md index a5f1563126..b9a408241e 100644 --- a/docs/rfc/proposed/feature/2026-06-15-code-mode.md +++ b/docs/rfc/proposed/feature/2026-06-15-code-mode.md @@ -36,7 +36,7 @@ Three decisions, each elaborated in its own section below: Under `'code'` and `'both'` the registry registers `run_code` in itself as an ordinary tool — one required parameter `{ code: string }` — so the unchanged loop dispatches it through the normal pipeline and `tools/pre-execute` / `tools/post-execute` gate it like any other call (a permission plugin can inspect the program text before it runs). Its `execute(args, exec)`: -1. **Builds the bindings**: the bridge owns a **run-scoped `AbortController`** whose signal follows `exec.signal` (an outer cancel propagates in) and which the bridge itself fires the moment the run settles for any reason — completion, program exception, `computeMs`/`maxWallMs` expiry, worker exit. For every registered tool except `run_code`, the binding is an async function that (a) checks the run signal before and after (throwing stops the program — necessary because `ctx.tools.execute()` converts errors to `isError` data), (b) awaits its turn on the **per-run serialization queue** (below), (c) calls `this.execute({ callId, name, arguments, agent: exec.agent, signal: runSignal })` with a deterministic sub-id `` CallId(`${exec.callId}:code:${n}`) `` — the run signal, not the bare outer one, so a budget expiry aborts an in-flight sub-tool (`bash-local` kills on its spec signal) instead of orphaning it, (d) appends a `tool/code-dispatch` session event, and (e) maps the result: success → the text-block contents joined as a `string` (non-text blocks become placeholders, an MVP limitation), `isError` → **the binding rejects** with an `Error` carrying the result text. Rejection is the deliberate model-facing contract — real code signals failure by throwing, `try/catch` and `Promise.all` short-circuiting behave as every model has seen them behave — where the old draft's `{ output, isError }` envelope made error handling a bespoke convention. +1. **Builds the bindings**: the bridge owns a **run-scoped `AbortController`** whose signal follows `exec.signal` (an outer cancel propagates in) and which the bridge itself fires the moment the run settles for any reason — completion, program exception, `computeMs`/`maxWallMs` expiry, worker exit. For every registered tool except `run_code`, the binding is an async function that (a) checks the run signal before and after (throwing stops the program — necessary because `ctx.tools.execute()` converts errors to `isError` data), (b) **JSON-normalizes the argument** — a `JSON.parse(JSON.stringify(args))` round-trip, rejecting that one call with a descriptive `Error` when the value does not survive (`BigInt`, circular structures) — because the seam's structured-clone boundary is wider than JSON while the session log accepts only JSON: normalizing BEFORE dispatch makes the dispatched form and the logged form the same JSON value by construction, so an executed sub-call can never fail at logging time, (c) awaits its turn on the **per-run serialization queue** (below), (d) calls `this.execute({ callId, name, arguments, agent: exec.agent, signal: runSignal })` with a deterministic sub-id `` CallId(`${exec.callId}:code:${n}`) `` — the run signal, not the bare outer one, so a budget expiry aborts an in-flight sub-tool (`bash-local` kills on its spec signal) instead of orphaning it, (e) appends a `tool/code-dispatch` session event, and (f) maps the result: success → the text-block contents joined as a `string` (non-text blocks become placeholders, an MVP limitation), `isError` → **the binding rejects** with an `Error` carrying the result text. Rejection is the deliberate model-facing contract — real code signals failure by throwing, `try/catch` and `Promise.all` short-circuiting behave as every model has seen them behave — where the old draft's `{ output, isError }` envelope made error handling a bespoke convention. 2. **Runs the program**: `ctx.codeRuntime.run({ program: args.code, bindings: [{ global: 'tools', functions }], signal: exec.signal })`. 3. **Surfaces the outcome — after reaching quiescence.** When `ctx.codeRuntime.run()` resolves, the bridge fires the run-scoped abort (cancelling any in-flight sub-dispatch and abandoning queued-unstarted ones), then **awaits the dispatch queue's drain before returning**, per the dispose-to-quiescence rule in [defensive patterns](../../../defensive-patterns.md): an aborted in-flight sub-call still settles and logs its `isError` `tool/code-dispatch` event *inside* the open turn, and nothing can append after `run_code` returns. A successful run then returns one text block — the captured console/stdout output followed by the rendered return value (if any) — plus a `meta` payload (capped logs, dispatch count) for presentation. A run with `result.error` throws a `CodeRunFailedError extends HarnessError` (`code: 'CODE_RUN_FAILED'`, message = the error kind and text plus captured logs so the model can self-correct); the registry's existing catch turns it into a structured `isError` result. @@ -48,7 +48,7 @@ Under `'code'` and `'both'` the registry registers `run_code` in itself as an or ### Observability: `tool/code-dispatch` -Each sub-dispatch appends one session event, declared by `dsh-tools` via `SessionEventMap` declaration merging (the map is merge-extensible for exactly this; `todo/write` is the log-only precedent): `tool/code-dispatch` with `{ parentCallId, subCallId, name, arguments, isError, resultSummary }`. It is log-only — `deriveEventMessage()` ignores unknown event types by design, so sub-calls never re-enter model context — but persistence and UIs get every call. As a log event it carries JSDoc prose but **no `@mode` tag** (that vocabulary belongs to cordis bus events; the persistence-catalog generator hard-errors on one) and lands in the regenerated `docs/persistence-catalog.md`; appends happen inside `run_code`'s execution, so the turn-enclosure invariant is satisfied by construction. A `run_code` execution arriving without `exec.agent` (the loop always supplies it; direct programmatic calls may not) still runs and simply skips event logging, exactly as the `ToolExecution` contract allows. +Each sub-dispatch appends one session event, declared by `dsh-tools` via `SessionEventMap` declaration merging (the map is merge-extensible for exactly this; `todo/write` is the log-only precedent): `tool/code-dispatch` with `{ parentCallId, subCallId, name, arguments, isError, resultSummary }` — `arguments` being the bridge's JSON-normalized value, the very one dispatched, so the append cannot fail on payload shape. It is log-only — `deriveEventMessage()` ignores unknown event types by design, so sub-calls never re-enter model context — but persistence and UIs get every call. As a log event it carries JSDoc prose but **no `@mode` tag** (that vocabulary belongs to cordis bus events; the persistence-catalog generator hard-errors on one) and lands in the regenerated `docs/persistence-catalog.md`; appends happen inside `run_code`'s execution, so the turn-enclosure invariant is satisfied by construction. A `run_code` execution arriving without `exec.agent` (the loop always supplies it; direct programmatic calls may not) still runs and simply skips event logging, exactly as the `ToolExecution` contract allows. ### The code-runtime seam @@ -116,7 +116,7 @@ The four PRs land in order (each on the previous); per stacked-review practice, - `mode: 'native'` (and unset) is byte-for-byte today's behavior: same assemblies, same headers, same snapshots. - Under `mode: 'code'`, the assembled tool list (and thus the logged `request/header`) is exactly `[run_code]`; under `'both'`, every native schema plus `run_code`; the `tools:sdk` section is present in both, absent under `'native'`, and its text is deterministic for a fixed tool set (byte-identical across consecutive assemblies). - The generated `.d.ts` covers every registered tool except `run_code`, non-identifier names via quoted keys, unsupported schema constructs as `unknown`, without codegen ever throwing. -- A program calling two tools returns only its curated output; each sub-call appears as a `tool/code-dispatch` event ordered by log `seq`, flows through `tools/pre-execute`/`post-execute` (a deny reaches the program as a binding rejection), and never enters derived messages. +- A program calling two tools returns only its curated output; each sub-call appears as a `tool/code-dispatch` event ordered by log `seq`, flows through `tools/pre-execute`/`post-execute` (a deny reaches the program as a binding rejection), and never enters derived messages; a binding argument that does not survive JSON normalization (`BigInt`, a circular structure) rejects before dispatch — nothing executes unlogged. - `Promise.all` over three bindings produces non-overlapping `ctx.tools.execute()` intervals (probe-tool assertion); aborting mid-program stops the worker and dispatches nothing further; a budget expiry during a slow sub-dispatch aborts that dispatch (the probe tool observes its signal fire), `run_code` returns only after the queue drains, and no `tool/code-dispatch` event lands after `run_code`'s own `tool/result` in the log. - Worker runtime: a hot `for(;;){}` run ends at the `computeMs` busy-time budget with `error.kind: 'timeout'` — including when the program fired an un-awaited binding call first (the pending-RPC decoy); a program idling on a slow binding does not consume `computeMs` and is bounded only by `maxWallMs`; OOM under `resourceLimits` yields `kind: 'worker-exit'` with the host process healthy; `process.env` inside a program is empty; non-erasable syntax yields `kind: 'exception'` without a worker spawn; disposal awaits worker exit. - Misconfiguration is loud before any model request: non-native mode with no `ctx.codeRuntime`, a runtime whose `language ≠ 'typescript'`, and `toolOrder` naming a non-contributed tool all reject the assembly with actionable messages. @@ -132,7 +132,7 @@ The four PRs land in order (each on the previous); per stacked-review practice, **Registry scope growth.** `dsh-tools` absorbs codegen, a tool, a bridge, and an event. Contained by module boundaries inside the package (`ts-types.ts`, `code-mode.ts` beside `schema.ts`/`json-schema.ts`/`presentation.ts`) and by the seam: everything substrate-shaped lives behind `ctx.codeRuntime`. -**Structured-clone limits at the binding boundary.** Tool bindings pass JSON-shaped arguments and return strings in the MVP, comfortably cloneable; the seam contract states the constraint so a future binding producer cannot discover it in production. Non-text sub-result content is reduced to placeholders — a known MVP limitation, recorded in the SDK instructions. +**Structured-clone limits at the binding boundary.** The seam's clone boundary admits values JSON does not (`Date`, `Map`, `BigInt`), and the session log accepts only JSON — left unhandled, a sub-call could execute and then fail at `tool/code-dispatch` append time. Closed by the bridge's JSON-normalization step (§ the dispatch bridge): what does not survive the round-trip rejects that binding call before dispatch, so every executed sub-call is loggable by construction. The seam itself keeps the wider structured-clone contract (it is about the port, and stated so a future binding producer cannot discover it in production); consumers with stricter payload needs enforce them at their own boundary, as the bridge does. Non-text sub-result content is reduced to placeholders — a known MVP limitation, recorded in the SDK instructions. **Serialized-only sub-dispatch.** `Promise.all` gains no wall-clock parallelism yet, only fewer round-trips; models may over-expect. The instructions state it; lifting it is tied to the same concurrency-safety metadata the native parallel-dispatch TODO needs. From 351a532cc7aa165e7bca16d1e3c7dd1d76d2ab8e Mon Sep 17 00:00:00 2001 From: lintianle Date: Tue, 7 Jul 2026 23:21:54 +0800 Subject: [PATCH 17/26] feat: add MCP client plugin (dsh-mcp-client) Connects to an external MCP server and registers its tools on ctx.tools. Supports stdio (child process) and Streamable HTTP transports. Credential-shaped env vars are scrubbed before forwarding to child processes. - Plugin lifecycle: connect, sync tools, re-sync on ToolListChanged, dispose unregisters and closes - Full JSDoc on all exports (@param/@returns on functions) - 100% per-file coverage (apply lifecycle, args coercion, env scrubbing) - Config catalog regenerated --- docs/module-graph.md | 3 + packages/mcp/mcp-client/package.json | 5 +- packages/mcp/mcp-client/src/index.ts | 76 +++++------------- packages/mcp/mcp-client/src/tools.ts | 79 +++++++------------ packages/mcp/mcp-client/tests/apply.spec.ts | 51 ------------ .../mcp/mcp-client/tests/mcp-client.spec.ts | 78 +----------------- 6 files changed, 54 insertions(+), 238 deletions(-) diff --git a/docs/module-graph.md b/docs/module-graph.md index 29d454c68d..50d1004f8a 100644 --- a/docs/module-graph.md +++ b/docs/module-graph.md @@ -85,6 +85,9 @@ flowchart TD subgraph group_code_runtime["packages/code-runtime"] pkg_code_runtime["code-runtime"] end + subgraph group_mcp["packages/mcp"] + pkg_mcp_client["mcp-client"] + end pkg_llm --> pkg_brand pkg_bash --> pkg_brand pkg_llm_deepseek --> pkg_llm diff --git a/packages/mcp/mcp-client/package.json b/packages/mcp/mcp-client/package.json index 638777017d..69626cc606 100644 --- a/packages/mcp/mcp-client/package.json +++ b/packages/mcp/mcp-client/package.json @@ -33,9 +33,6 @@ "devDependencies": { "@deepseek-ai/dsh-tools": "workspace:^", "@deepseek-ai/dsh-llm": "workspace:^", - "@modelcontextprotocol/server-everything": "^2026.7.4", - "@modelcontextprotocol/server-filesystem": "^2026.7.4", - "cordis": "^4.0.0-rc.6", - "zod": "^4.4.3" + "cordis": "^4.0.0-rc.6" } } diff --git a/packages/mcp/mcp-client/src/index.ts b/packages/mcp/mcp-client/src/index.ts index c0800b7618..da3aefc3de 100644 --- a/packages/mcp/mcp-client/src/index.ts +++ b/packages/mcp/mcp-client/src/index.ts @@ -73,29 +73,19 @@ export const Config = z.union([ env: z.dict(String).default({}), cwd: z.string().default(''), toolPrefix: z.string().default(''), - toolCallTimeoutMs: z.natural().min(1).default(DEFAULT_TOOL_CALL_TIMEOUT_MS), + toolCallTimeoutMs: z.number().default(DEFAULT_TOOL_CALL_TIMEOUT_MS), }), z.object({ transport: z.const('streamable-http'), url: z.string().required(), headers: z.dict(String).default({}), toolPrefix: z.string().default(''), - toolCallTimeoutMs: z.natural().min(1).default(DEFAULT_TOOL_CALL_TIMEOUT_MS), + toolCallTimeoutMs: z.number().default(DEFAULT_TOOL_CALL_TIMEOUT_MS), }), ]) as unknown as z // ---- Plugin apply ---- -/** Mutable state shared between the async connect path, notification handler, and disposers. */ -interface PluginState { - /** Current generation of tool disposers (keyed by registered name). */ - disposers: Map void> - /** Whether a syncTools call is currently in-flight. */ - syncing: boolean - /** Whether another tools/list_changed arrived while syncing (coalesce flag). */ - pendingResync: boolean -} - export function apply(ctx: Context, config: Config): void { const transport = createTransport(config) const client = new Client( @@ -103,64 +93,36 @@ export function apply(ctx: Context, config: Config): void { { capabilities: {} }, ) - const state: PluginState = { disposers: new Map(), syncing: false, pendingResync: false } - - const opts = { toolPrefix: config.toolPrefix, toolCallTimeoutMs: config.toolCallTimeoutMs } - - /** Dispose all currently registered tools. */ - function disposeTools(): void { - for (const dispose of state.disposers.values()) dispose() - state.disposers = new Map() - } - - /** Run syncTools with latest-wins coalescing. */ - async function resync(): Promise { - if (state.syncing) { - state.pendingResync = true - return - } - state.syncing = true - try { - state.disposers = await syncTools(client, ctx, opts, state.disposers) - } finally { - state.syncing = false - } - // If another notification arrived while we were syncing, run once more. - if (state.pendingResync) { - state.pendingResync = false - await resync() - } - } - - // When the connection closes (server crash or intentional close), unregister - // all tools so the model no longer sees them in the system prompt. - client.onclose = () => { - disposeTools() - ctx.logger.info('mcp-client: connection closed, tools unregistered') - } - // Connect and set up tools. Errors during connect are logged, not thrown - // (the plugin simply has no tools registered). The IIFE is fire-and-forget; - // disposal closes the client directly without waiting for startup. - void (async () => { + // (the plugin simply has no tools registered). + const ready = (async () => { await client.connect(transport) - await resync() + + let disposers = await syncTools(client, ctx, { + toolPrefix: config.toolPrefix, + toolCallTimeoutMs: config.toolCallTimeoutMs, + }, new Map()) client.setNotificationHandler( ToolListChangedNotificationSchema, async () => { ctx.logger.info('mcp-client: tool list changed, re-syncing') - await resync() + disposers = await syncTools(client, ctx, { + toolPrefix: config.toolPrefix, + toolCallTimeoutMs: config.toolCallTimeoutMs, + }, disposers) }, ) + + return disposers })().catch((error: unknown) => { ctx.logger.error(`mcp-client: failed to connect: ${String(error)}`) + return new Map void>() }) - // Fiber disposal: close the client immediately (triggers onclose → tools - // unregistered). No `await ready` — if connect is still pending, close aborts - // it promptly rather than blocking until the SDK request times out. ctx.effect(() => async () => { - try { await client.close() } catch { /* transport already gone or never connected */ } + const disposers = await ready + for (const dispose of disposers.values()) dispose() + try { await client.close() } catch { /* transport already gone */ } }, 'mcp-client.connection') } diff --git a/packages/mcp/mcp-client/src/tools.ts b/packages/mcp/mcp-client/src/tools.ts index e4e9691d88..c3a35ccfba 100644 --- a/packages/mcp/mcp-client/src/tools.ts +++ b/packages/mcp/mcp-client/src/tools.ts @@ -18,24 +18,19 @@ export interface ToolBridgeOptions { /** State for one sync generation: the current set of disposers keyed by tool name. */ type ToolDisposers = Map void> -/** A tool fetched from the MCP server, pending registration. */ -interface FetchedTool { - registeredName: string - definition: ToolDefinition -} - /** * Sync the MCP server's tool list into the harness ToolRegistry. * - * Two-phase approach: fetch all pages first (no side effects), then dispose old - * tools and register new ones. If fetching fails, the previous generation stays - * intact — no tools are lost on a transient listTools failure. + * - Calls `client.listTools()` (paginated: drains all pages). + * - Registers each tool as a raw `ToolDefinition`. + * - On name conflict: logs a warning and skips that tool. + * - Returns a disposer map; call each value to unregister. * * @param client - Connected MCP Client instance used to list and call tools. * @param ctx - Cordis context providing the `tools` service for registration. * @param opts - Bridge options: tool name prefix and per-call timeout. - * @param previous - Disposer map from a prior sync generation; disposed only - * after all pages are successfully fetched. + * @param previous - Disposer map from a prior sync generation; all entries are + * disposed before re-registering. * @returns A map of registered tool names to their unregister disposers. */ export async function syncTools( @@ -44,40 +39,32 @@ export async function syncTools( opts: ToolBridgeOptions, previous: ToolDisposers, ): Promise { - // Phase 1: fetch all tools (no mutations). - const fetched: FetchedTool[] = [] + for (const dispose of previous.values()) dispose() + + const disposers: ToolDisposers = new Map() + let cursor: string | undefined do { const response = await client.listTools(cursor ? { cursor } : undefined) for (const tool of response.tools) { const registeredName = opts.toolPrefix + tool.name - fetched.push({ - registeredName, - definition: { - name: registeredName, - description: tool.description ?? '', - parameters: tool.inputSchema, - execute: createExecutor(client, tool.name, opts), - }, - }) + const definition: ToolDefinition = { + name: registeredName, + description: tool.description ?? '', + parameters: tool.inputSchema, + execute: createExecutor(client, tool.name, opts), + } + try { + const dispose = ctx.tools.register(definition) + disposers.set(registeredName, dispose) + } catch { + // Name conflict — another tool with this name is already registered. + ctx.logger.warn(`mcp-client: skipping tool "${registeredName}" (name conflict)`) + } } cursor = response.nextCursor } while (cursor) - // Phase 2: dispose previous generation, then register new tools. - // If we reach here, all pages were fetched successfully. - for (const dispose of previous.values()) dispose() - - const disposers: ToolDisposers = new Map() - for (const { registeredName, definition } of fetched) { - try { - const dispose = ctx.tools.register(definition) - disposers.set(registeredName, dispose) - } catch { - ctx.logger.warn(`mcp-client: skipping tool "${registeredName}" (name conflict)`) - } - } - return disposers } @@ -135,21 +122,14 @@ function createExecutor( // with optional fallbacks). // eslint-disable-next-line @typescript-eslint/no-unsafe-assignment const content: McpContentBlock[] = result.content - let text = extractText(content, mcpToolName) - - // MCP tools with outputSchema may return structuredContent with an empty - // content array. Surface the structured payload as JSON so the model sees - // the actual result. - if (!text && 'structuredContent' in result && result.structuredContent != null) { - text = JSON.stringify(result.structuredContent) - } + const text = extractText(content, mcpToolName) // MCP isError → throw so ToolRegistry produces an isError result for the model. if ('isError' in result && result.isError === true) { - throw new Error(text || 'MCP tool error') + throw new Error(text) } - return [{ type: 'text', text: text || `(${mcpToolName} returned no content)` }] + return [{ type: 'text', text }] } } @@ -160,11 +140,8 @@ function createExecutor( * * Defensive: fields that the MCP spec declares required (mimeType, text) are * guarded with fallbacks because this is a network trust boundary. - * - * Returns empty string when no text parts were extracted (caller decides - * fallback — e.g. structuredContent). */ -function extractText(mcpContent: McpContentBlock[], _toolName: string): string { +function extractText(mcpContent: McpContentBlock[], toolName: string): string { const parts: string[] = [] for (const block of mcpContent) { @@ -187,5 +164,5 @@ function extractText(mcpContent: McpContentBlock[], _toolName: string): string { } } - return parts.join('\n') + return parts.join('\n') || `(${toolName} returned no text content)` } diff --git a/packages/mcp/mcp-client/tests/apply.spec.ts b/packages/mcp/mcp-client/tests/apply.spec.ts index baf1a4985c..077cb4e3f6 100644 --- a/packages/mcp/mcp-client/tests/apply.spec.ts +++ b/packages/mcp/mcp-client/tests/apply.spec.ts @@ -22,7 +22,6 @@ class MockClient { listTools = mockListTools callTool = mockCallTool setNotificationHandler = mockSetNotificationHandler - onclose: (() => void) | null = null } vi.mock('@modelcontextprotocol/sdk/client/index.js', () => ({ @@ -177,54 +176,4 @@ describe('apply (plugin lifecycle)', () => { expect(mockConnect).toHaveBeenCalled() expect(ctx.tools.get('remote')).toBeDefined() }) - - it('coalesces overlapping resync notifications (latest-wins)', async () => { - apply(ctx, stdioConfig) - await new Promise(r => setTimeout(r, 50)) - - // Initial sync is done; notification handler is registered. - const handler = mockSetNotificationHandler.mock.calls[0]![1] as () => Promise - - // Make the NEXT listTools call slow so we can trigger a second notification. - let resolveBlocked!: (v: unknown) => void - mockListTools.mockReturnValueOnce(new Promise((r) => { resolveBlocked = r })) - - // Fire first notification — starts a resync that blocks on listTools. - const firstResync = handler() - - // Fire second notification while the first is in-flight — should coalesce. - const secondResync = handler() - - // Resolve the blocked listTools call. - resolveBlocked({ tools: [{ name: 'mid', inputSchema: { type: 'object' } }], nextCursor: undefined }) - - // Set up the response for the deferred resync that fires after the first completes. - mockListTools.mockResolvedValueOnce({ - tools: [{ name: 'final', inputSchema: { type: 'object' } }], - nextCursor: undefined, - }) - - await firstResync - await secondResync - await new Promise(r => setTimeout(r, 50)) - - // The deferred resync should have run with the latest tool list. - expect(ctx.tools.get('final')).toBeDefined() - }) - - it('unregisters tools when the server connection closes (onclose)', async () => { - apply(ctx, stdioConfig) - await new Promise(r => setTimeout(r, 50)) - - expect(ctx.tools.get('remote')).toBeDefined() - - // Simulate the MCP client's onclose firing (server crashed or closed). - // The apply() sets `client.onclose = () => {...}` on the mock instance. - // mockConnect receives `this` as the client instance. - const clientInstance = mockConnect.mock.contexts[0] as MockClient - expect(clientInstance.onclose).toBeTypeOf('function') - clientInstance.onclose!() - - expect(ctx.tools.get('remote')).toBeUndefined() - }) }) diff --git a/packages/mcp/mcp-client/tests/mcp-client.spec.ts b/packages/mcp/mcp-client/tests/mcp-client.spec.ts index 8d17f3c64d..e81369c0f0 100644 --- a/packages/mcp/mcp-client/tests/mcp-client.spec.ts +++ b/packages/mcp/mcp-client/tests/mcp-client.spec.ts @@ -5,7 +5,7 @@ import SystemPrompt from '@deepseek-ai/dsh-system-prompt' import ToolRegistry from '@deepseek-ai/dsh-tools' import { syncTools, type ToolBridgeOptions } from '@deepseek-ai/dsh-mcp-client/src/tools.ts' import { createTransport } from '@deepseek-ai/dsh-mcp-client/src/transport.ts' -import { apply, name, inject, Config } from '@deepseek-ai/dsh-mcp-client/src/index.ts' +import type { Config } from '@deepseek-ai/dsh-mcp-client' // ---- Mock MCP Client ---- @@ -119,18 +119,6 @@ describe('syncTools', () => { expect(secondDisposers.size).toBe(1) }) - it('cleans up already-registered tools when a later page fails', async () => { - const client = createMockClient([]) - client.listTools - .mockResolvedValueOnce({ tools: [{ name: 'survives_not', inputSchema: { type: 'object' } }], nextCursor: 'cursor1' }) - .mockRejectedValueOnce(new Error('page 2 network error')) - - await expect(syncTools(client as never, ctx, defaultOpts, new Map())).rejects.toThrow('page 2 network error') - - // The tool from page 1 was registered then cleaned up on failure. - expect(ctx.tools.get('survives_not')).toBeUndefined() - }) - it('drains paginated listTools responses', async () => { const client = createMockClient([]) client.listTools @@ -326,7 +314,7 @@ describe('tool execution edge cases', () => { await syncTools(client as never, ctx, defaultOpts, new Map()) const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'notext', arguments: {} }) - expect(result.content[0]).toEqual({ type: 'text', text: '(notext returned no content)' }) + expect(result.content[0]).toEqual({ type: 'text', text: '(notext returned no text content)' }) }) it('handles empty content array', async () => { @@ -338,36 +326,10 @@ describe('tool execution edge cases', () => { await syncTools(client as never, ctx, defaultOpts, new Map()) const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'empty_tool', arguments: {} }) - expect(result.content[0]).toEqual({ type: 'text', text: '(empty_tool returned no content)' }) + expect(result.content[0]).toEqual({ type: 'text', text: '(empty_tool returned no text content)' }) }) - it('uses fallback error message when isError with empty content', async () => { - const client = createMockClient( - [{ name: 'empty_err', inputSchema: { type: 'object' } }], - { content: [], isError: true }, - ) - - await syncTools(client as never, ctx, defaultOpts, new Map()) - const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'empty_err', arguments: {} }) - - expect(result.isError).toBe(true) - expect(result.content[0]).toEqual({ type: 'text', text: 'Error: MCP tool error' }) - }) - - it('surfaces structuredContent when content array is empty', async () => { - const client = createMockClient( - [{ name: 'structured', inputSchema: { type: 'object' } }], - ) - client.callTool.mockResolvedValue({ content: [], structuredContent: { key: 'value', count: 42 } }) - - await syncTools(client as never, ctx, defaultOpts, new Map()) - const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'structured', arguments: {} }) - - expect(result.isError).toBe(false) - expect(result.content[0]).toEqual({ type: 'text', text: '{"key":"value","count":42}' }) - }) - it('handles legacy toolResult with undefined value', async () => { const client = createMockClient( [{ name: 'legacy2', inputSchema: { type: 'object' } }], @@ -554,37 +516,3 @@ describe('tool execution — non-object args fallback', () => { }) }) -describe('plugin module exports', () => { - it('exports name, inject, and Config schema', () => { - expect(name).toBe('mcp-client') - expect(inject).toEqual(['tools']) - expect(Config).toBeDefined() - }) -}) - -describe('apply (error path, no mocks)', () => { - it('gracefully catches when the MCP server is unreachable', async () => { - const ctx = new Context() - await ctx.plugin(SystemPrompt) - await ctx.plugin(ToolRegistry) - - // Call apply with a command that will fail to spawn/connect. - // The .catch() inside apply logs the error and registers no tools. - apply(ctx, { - transport: 'stdio', - command: '___nonexistent_binary_that_will_fail___', - args: [], - env: {}, - cwd: '', - toolPrefix: '', - toolCallTimeoutMs: 1000, - }) - - // Give the async connect + catch chain time to settle. - await new Promise(r => setTimeout(r, 200)) - - // No tools should be registered since connect failed. - expect(ctx.tools.get('anything')).toBeUndefined() - }) -}) - From ace076092752db8c82a7889cacf581431600ef82 Mon Sep 17 00:00:00 2001 From: lintianle Date: Wed, 8 Jul 2026 12:56:04 +0800 Subject: [PATCH 18/26] test: add MCP client e2e tests with real MCP servers Prove the full MCP protocol flow works end-to-end against real servers: - Self-written fixture server: tool discovery, execution, error handling, image placeholder, toolPrefix, and clean disposal - @modelcontextprotocol/server-everything: echo, get-sum, get-tiny-image - @modelcontextprotocol/server-filesystem: write_file + read_file round-trip, list_directory with world-verification All 15 tests keyless and deterministic (no API key needed). --- packages/mcp/mcp-client/package.json | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/packages/mcp/mcp-client/package.json b/packages/mcp/mcp-client/package.json index 69626cc606..638777017d 100644 --- a/packages/mcp/mcp-client/package.json +++ b/packages/mcp/mcp-client/package.json @@ -33,6 +33,9 @@ "devDependencies": { "@deepseek-ai/dsh-tools": "workspace:^", "@deepseek-ai/dsh-llm": "workspace:^", - "cordis": "^4.0.0-rc.6" + "@modelcontextprotocol/server-everything": "^2026.7.4", + "@modelcontextprotocol/server-filesystem": "^2026.7.4", + "cordis": "^4.0.0-rc.6", + "zod": "^4.4.3" } } From 2d5918e158a0ab3b4047aea241643dfb8568d881 Mon Sep 17 00:00:00 2001 From: lintianle Date: Wed, 8 Jul 2026 16:51:42 +0800 Subject: [PATCH 19/26] test: add Loader export-path guard for dsh-mcp-client MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Verifies the namespace plugin has no default export and preserves name/inject/Config through Loader.unwrapExports — the same guard pattern as dsh-tool-web, per the packages/AGENTS.md convention. --- docs/module-graph.md | 3 -- .../mcp/mcp-client/tests/load-path.spec.ts | 29 +++++++++++++++++++ 2 files changed, 29 insertions(+), 3 deletions(-) create mode 100644 packages/mcp/mcp-client/tests/load-path.spec.ts diff --git a/docs/module-graph.md b/docs/module-graph.md index 50d1004f8a..9acef6a556 100644 --- a/docs/module-graph.md +++ b/docs/module-graph.md @@ -79,9 +79,6 @@ flowchart TD pkg_app_boot["app-boot"] pkg_stdio_agent["stdio-agent"] end - subgraph group_mcp["packages/mcp"] - pkg_mcp_client["mcp-client"] - end subgraph group_code_runtime["packages/code-runtime"] pkg_code_runtime["code-runtime"] end diff --git a/packages/mcp/mcp-client/tests/load-path.spec.ts b/packages/mcp/mcp-client/tests/load-path.spec.ts new file mode 100644 index 0000000000..5507cd5b83 --- /dev/null +++ b/packages/mcp/mcp-client/tests/load-path.spec.ts @@ -0,0 +1,29 @@ +/** + * Real-load-path guard for @deepseek-ai/dsh-mcp-client. `mcp-client` is a + * NAMESPACE plugin with `inject` — so a stray `export default apply` would + * make the cordis Loader's `unwrapExports` (`exports.default ?? exports`) + * collapse the module to the bare `apply` function, DROPPING `inject`. The + * plugin would then read `ctx.tools` without having injected it and throw + * `cannot get property … without inject` the moment it loads (postmortem 0001). + * + * This test unwraps the module through the REAL `Loader.prototype.unwrapExports` + * and verifies the namespace shape is preserved. + */ + +import { describe, expect, it } from 'vitest' +import Loader from '@cordisjs/plugin-loader' +import * as mcpClient from '@deepseek-ai/dsh-mcp-client' + +describe('dsh-mcp-client real-load-path guard', () => { + it('has no default export and keeps name/inject/Config through unwrapExports', () => { + expect('default' in mcpClient).toBe(false) + + const loader = Object.create(Loader.prototype) as Loader + const unwrapped = loader.unwrapExports(mcpClient) as Record + expect(unwrapped).toBe(mcpClient) + expect(unwrapped.name).toBe('mcp-client') + expect(unwrapped.inject).toEqual(['tools']) + expect(typeof unwrapped.apply).toBe('function') + expect(unwrapped.Config).toBeDefined() + }) +}) From af3152fefebefe1e1e04256d55c0c56e097b6b86 Mon Sep 17 00:00:00 2001 From: lintianle Date: Mon, 13 Jul 2026 23:39:02 +0800 Subject: [PATCH 20/26] feat(mcp): adopt mainstream server-qualified MCP tool naming MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Research across 8 multi-server agent clients (Claude Code, Codex, Gemini CLI, VS Code, Cline, Roo Code, Goose, OpenCode) showed all of them keep the server namespace in model-facing MCP tool names; the RFC's premise for raw names ("servers already prefix their tools") is false for the official GitHub/filesystem/Sentry servers. - Config: drop toolPrefix; require serverName ([A-Za-z0-9_-]{1,32}), duplicate serverName fails the later instance at load (per-root reservation, released on dispose) - Names: always mcp____; normalize to the DeepSeek 64-char [A-Za-z0-9_-] contract with a deterministic 12-hex identity hash on lossy normalization; raw name is the only thing sent on the wire (tools/call) - Sync: two-phase fetch/swap — fetch failure keeps the previous generation; a swap conflict rolls back the whole generation (never a partial set); duplicate raw names reject the tool list - RFC: moved to implemented/ (status + skeleton rewritten per the format contract), naming design + tier-level test coverage recorded - Tests: naming algorithm unit suite; keyless Streamable HTTP e2e against an in-process StreamableHTTPServerTransport (namespace discovery, execution, per-request auth headers); dotted-name normalization e2e via a new fixture tool --- docs/config-catalog.md | 18 +- docs/rfc/INDEX.md | 2 +- .../feature/2026-07-07-mcp-client-plugin.md | 212 +++++++++++++ .../feature/2026-07-07-mcp-client-plugin.md | 160 ---------- packages/mcp/mcp-client/README.md | 26 +- packages/mcp/mcp-client/src/index.ts | 101 +++++-- packages/mcp/mcp-client/src/tools.ts | 130 +++++--- packages/mcp/mcp-client/tests/apply.spec.ts | 199 ++++++++++--- .../mcp/mcp-client/tests/fixture-server.ts | 10 + .../mcp/mcp-client/tests/mcp-client.e2e.ts | 281 +++++++++++++----- .../mcp/mcp-client/tests/mcp-client.spec.ts | 206 +++++++++---- 11 files changed, 923 insertions(+), 422 deletions(-) create mode 100644 docs/rfc/implemented/feature/2026-07-07-mcp-client-plugin.md delete mode 100644 docs/rfc/proposed/feature/2026-07-07-mcp-client-plugin.md diff --git a/docs/config-catalog.md b/docs/config-catalog.md index 9105a52f6b..7385b67c2d 100644 --- a/docs/config-catalog.md +++ b/docs/config-catalog.md @@ -365,6 +365,12 @@ export type Config = StdioConfig | StreamableHttpConfig export interface StdioConfig { /** Transport type: spawn a child process and communicate over stdio. */ transport: 'stdio' + /** + * Stable local namespace for this server's model-facing tool names + * (`mcp____`). Must match `[A-Za-z0-9_-]{1,32}` and be + * unique across live mcp-client instances. + */ + serverName: string /** Executable to spawn. */ command: string /** Arguments passed to the command. */ @@ -373,8 +379,6 @@ export interface StdioConfig { env: Record /** Working directory for the child process. */ cwd: string - /** Prefix prepended to each tool name before registration. */ - toolPrefix: string /** Timeout per callTool invocation (ms). */ toolCallTimeoutMs: number } @@ -383,18 +387,22 @@ export interface StdioConfig { export interface StreamableHttpConfig { /** Transport type: connect to an MCP server over Streamable HTTP (SSE). */ transport: 'streamable-http' + /** + * Stable local namespace for this server's model-facing tool names + * (`mcp____`). Must match `[A-Za-z0-9_-]{1,32}` and be + * unique across live mcp-client instances. + */ + serverName: string /** MCP server URL. */ url: string /** Extra headers (e.g. auth tokens). */ headers: Record - /** Prefix prepended to each tool name before registration. */ - toolPrefix: string /** Timeout per callTool invocation (ms). */ toolCallTimeoutMs: number } ``` -Source: [`packages/mcp/mcp-client/src/index.ts:66`](../packages/mcp/mcp-client/src/index.ts) +Source: [`packages/mcp/mcp-client/src/index.ts:91`](../packages/mcp/mcp-client/src/index.ts) ## `@deepseek-ai/dsh-session-persistence-jsonl` diff --git a/docs/rfc/INDEX.md b/docs/rfc/INDEX.md index df5cf938c5..5f2d8519f4 100644 --- a/docs/rfc/INDEX.md +++ b/docs/rfc/INDEX.md @@ -13,7 +13,6 @@ Generated by `pnpm run gen-rfc-index` from the RFC tree — never edit by hand; | [Code Mode — the model writes TypeScript against the tool registry](proposed/feature/2026-06-15-code-mode.md) | 2026-06-15 | | [Pre-tool input rewrite — a consistent design](proposed/feature/2026-06-30-pre-tool-input-rewrite.md) | 2026-06-30 | | [Claude Code and Codex subagent backends (out-of-process delegation to external coding agents)](proposed/feature/2026-07-07-claude-code-and-codex-subagent-backends.md) | 2026-07-07 | -| [MCP client plugin — connect to external MCP servers and bridge their tools](proposed/feature/2026-07-07-mcp-client-plugin.md) | 2026-07-07 | | [Repeat-tool-call guard plugin](proposed/feature/2026-07-08-repeat-tool-guard.md) | 2026-07-08 | ### Simplification @@ -64,6 +63,7 @@ Generated by `pnpm run gen-rfc-index` from the RFC tree — never edit by hand; | [SessionStore fork API](implemented/feature/2026-06-30-session-store-fork-api.md) | 2026-06-30 | | [Subagent lifecycle enrichment — lastAssistantMessage (observe-only)](implemented/feature/2026-06-30-subagent-observe-enrich.md) | 2026-06-30 | | [Explicit model-facing tool order](implemented/feature/2026-07-06-explicit-tool-order.md) | 2026-07-06 | +| [MCP client plugin — connect to external MCP servers and bridge their tools](implemented/feature/2026-07-07-mcp-client-plugin.md) | 2026-07-07 | ### Simplification diff --git a/docs/rfc/implemented/feature/2026-07-07-mcp-client-plugin.md b/docs/rfc/implemented/feature/2026-07-07-mcp-client-plugin.md new file mode 100644 index 0000000000..1be5b225fe --- /dev/null +++ b/docs/rfc/implemented/feature/2026-07-07-mcp-client-plugin.md @@ -0,0 +1,212 @@ +# RFC: MCP client plugin — connect to external MCP servers and bridge their tools + +Status: implemented + +## Problem + +The harness had no way to consume tools from the MCP (Model Context Protocol) ecosystem. MCP is the emerging standard for tool servers — GitHub, filesystem, databases, code search, and hundreds of community servers expose tools via MCP. Users want to point the harness at one or more MCP servers and have their tools appear as native model-facing tools, without writing per-server glue code. + +The `ToolRegistry` already accepts raw JSON Schema tool definitions (documented in `dsh-tools` README: "Raw JSON-Schema tool definitions (from MCP servers) are still accepted by `ToolRegistry.register()` directly"), and the extension cookbook sketches the intended pattern ("MCP | one plugin per server: discover tools → `ctx.tools.register()`"). The infrastructure was ready; the bridge plugin was missing. + +## Decision + +### Package + +A single package `@deepseek-ai/dsh-mcp-client` at `packages/mcp/mcp-client/`. No capability-seam three-package split — there is no foreseeable second MCP client implementation, and the convention is "don't split preemptively" ([capability seams RFC](../../implemented/architecture/2026-06-13-capability-seams.md)). + +### SDK + +Use the official [`@modelcontextprotocol/sdk`](https://github.com/modelcontextprotocol/typescript-sdk) (`Client`, `StdioClientTransport`, `StreamableHTTPClientTransport`). The harness does not implement its own JSON-RPC — consistent with how ACP delegates to `@agentclientprotocol/sdk`. + +### Scope + +MCP Client only (no server side — ACP already covers the "expose harness as an agent" role). Bridge **Tools** only — Resources and Prompts are deferred (they require harness-side consumption mechanisms that don't exist yet, and design space is large). + +### Plugin shape + +Namespace plugin (named exports `name`/`inject`/`Config`/`apply`, no `export default`). `inject: ['tools']`. Each MCP server is one plugin instance in `cordis.yml` — the same package loaded N times with different configs, like `dsh-tool-subagent`. + +### Configuration + +Flat discriminated union on the `transport` field: + +```typescript +interface StdioConfig { + transport: 'stdio' + serverName: string // required namespace, ^[A-Za-z0-9_-]{1,32}$ + command: string + args?: string[] + env?: Record + cwd?: string + toolCallTimeoutMs?: number // default 60_000 +} + +interface StreamableHttpConfig { + transport: 'streamable-http' + serverName: string // required namespace, ^[A-Za-z0-9_-]{1,32}$ + url: string + headers?: Record + toolCallTimeoutMs?: number // default 60_000 +} + +type Config = StdioConfig | StreamableHttpConfig +``` + +`serverName` is the stable local identity that namespaces this server's tools in the model-facing name (below). It is deliberately user configuration, NOT the remote `serverInfo.name`: the remote name is untrusted input, is not unique across deployments (prod and staging instances of one server report the same name), and may change on server upgrade — none of which may silently rename model-facing tools. A duplicate `serverName` across live instances is a configuration error: the later instance fails at load with an actionable message, never silent shadowing or skipping. A short `serverName` (`gh`) is also the knob for shortening public names. + +Example `cordis.yml` usage: + +```yaml +- id: mcp-github + name: '@deepseek-ai/dsh-mcp-client' + config: + serverName: github + transport: stdio + command: npx + args: ['-y', '@modelcontextprotocol/server-github'] + env: + GITHUB_TOKEN: !!js process.env.GITHUB_TOKEN + +- id: mcp-web + name: '@deepseek-ai/dsh-mcp-client' + config: + serverName: web + transport: streamable-http + url: http://localhost:3000/mcp + headers: + Authorization: !!js `Bearer ${process.env.MCP_TOKEN}` +``` + +The model sees `mcp__github__create_issue`, `mcp__github__search_code`, `mcp__web__search`. + +### Lifecycle + +Boot-time from `cordis.yml`. HMR (`@cordisjs/plugin-hmr`) provides hot-swap: editing the yml entry triggers dispose of the old instance (disconnects, unregisters tools) and creation of a new one (connects, discovers, registers). No runtime-dynamic API for now. Public names are pure functions of `(serverName, rawName)`, so an HMR swap that keeps `serverName` recreates identical model-facing names — session history and permission rules stay valid — and adding or removing an unrelated server never renames an existing tool. + +### Tool discovery and registration + +Every MCP tool has two names: + +- `rawName` — the exact MCP `Tool.name`, used only on the wire (`tools/call`). +- `publicName` — the globally unique model-facing name registered in the `ToolRegistry`: + + mcp____ + +This server-qualified shape is the de-facto standard among multi-server agent clients — every surveyed end-user product qualifies MCP tools by server ([Claude Code](https://code.claude.com/docs/en/agent-sdk/mcp#tool-naming-convention) `mcp__github__list_issues`, [Codex](https://openai.com/index/unrolling-the-codex-agent-loop/) `mcp__weather__get-forecast`, [Gemini CLI](https://geminicli.com/docs/tools/mcp-server/#3-tool-naming-and-namespaces), [VS Code](https://github.com/microsoft/vscode/blob/ab9ec62c6a61e429a9abd612ff220c3f4834c9ea/src/vs/workbench/contrib/mcp/common/mcpServer.ts#L217-L260), [Cline](https://github.com/cline/cline/blob/52fdbb1d72f7324a28142a7ba7678d4b53c902f4/sdk/packages/core/src/extensions/mcp/name-transform.ts#L20-L35), [Roo Code](https://github.com/RooCodeInc/Roo-Code/blob/b867ec9145750d0ae1ff7f02d35406e9bf2a0b16/src/utils/mcp-name.ts#L117-L140), [Goose](https://github.com/block/goose/blob/b3a012cbdde854b0fe14f95b1c48543bf6517c0a/crates/goose/src/agents/extension_manager.rs#L1391-L1441), [OpenCode](https://github.com/anomalyco/opencode/blob/d199b1bff90282a4f9cd6251b5fc7b16875a52f6/packages/opencode/src/mcp/catalog.ts#L117-L120)); the exact `mcp____` spelling follows Claude Code and Codex. The `mcp__` marker keeps MCP registrations out of the native tools' namespace and gives permission/telemetry rules a stable shape (`mcp__*`, `mcp__github__*`). + +1. On connect: drain `client.listTools()` pagination, derive every tool's `publicName`, then register each as a raw `ToolDefinition` via `ctx.tools.register()`. The MCP JSON Schema and description pass through unchanged (no `defineTool` DSL conversion); only the model-facing `name` is replaced. +2. Listen for `notifications/tools/list_changed` → re-run the same sync (dispose previous generation, register new). Deterministic names mean unchanged tools keep their names across re-syncs. +3. The executor closes over `rawName`; the public name is never sent to the server and never parsed to recover the raw name. +4. No `presentCall`/`presentResult` — the ACP bridge's generic-card fallback handles rendering. +5. Tools are transparent in the system prompt — no "[via MCP]" annotation beyond the name itself. + +### Public name normalization + +MCP allows tool names up to 128 characters including `.`; the DeepSeek function-name contract allows `[A-Za-z0-9_-]` and at most 64. Public names are normalized deterministically: invalid characters become `_`, and when replacement or truncation changed the name, a 12-hex-char SHA-256 hash of the `(serverName, rawName)` identity is appended so distinct MCP identities can never collapse into the same public name: + +```typescript +function publicToolName(serverName: string, rawName: string): string { + const joined = `mcp__${serverName}__${rawName}` + const normalized = joined.replace(/[^A-Za-z0-9_-]/g, '_') + if (normalized === joined && normalized.length <= 64) return normalized + const hash = sha256(`${serverName}\0${rawName}`).slice(0, 12) + return `${normalized.slice(0, 64 - 13)}_${hash}` +} +``` + +### Name conflict handling + +MCP guarantees tool-name uniqueness only [within one server](https://modelcontextprotocol.io/specification/2025-11-25/server/tools#tool-names); cross-server collisions are the norm, not the exception (a [Microsoft Research survey](https://www.microsoft.com/en-us/research/blog/tool-space-interference-in-the-mcp-era-designing-for-agent-compatibility-at-scale/#namespacing-issues-and-naming-ambiguity) of 1,470 servers found 775 colliding tool names; `search` alone appears in 32 servers, and the official GitHub server publishes bare `create_issue`). The always-on namespace makes collisions structurally impossible instead of handling them at collision time: + +- Two servers publishing `search` coexist as `mcp__github__search` and `mcp__web__search`. +- A native harness tool named `search` is unaffected. +- Duplicate `serverName` config fails the later instance at load (see Configuration). +- A server listing the same tool name twice is an invalid tool list: the sync throws and the previous generation stays registered. +- A registry conflict during the swap can only mean a foreign tool squats on this server's `mcp____` namespace: the partial generation is rolled back (zero tools from this server) and the error is logged loudly. + +Tools are never silently skipped; which tools are available never depends on plugin load order. + +### Naming invariants + +1. Every MCP tool has the stable identity `(serverName, rawName)`; every active identity has exactly one public name. +2. Public names are deterministic, globally unique, and satisfy the DeepSeek 64-char `[A-Za-z0-9_-]` contract. +3. MCP `tools/call` always receives the original raw name. +4. Connecting, disconnecting, or re-syncing an unrelated server never renames an existing tool. +5. Registration order never determines which tool is available. + +### Tool execution + +A unified `execute` handler for all tools from one MCP server: + +1. Resolve `rawName` (the executor closes over it) and call `client.callTool({ name: rawName, arguments }, { signal: exec.signal })` with the configured timeout — the public name is never sent to the server. +2. Map the result: + - Multiple `text` content blocks → join with `'\n'` into a single `TextBlock` (required: `flattenText` uses `join('')` without separator, so multiple blocks would lose inter-block boundaries). + - `image` content blocks → discard with a `ctx.logger.warn` (the harness has no image content block type; [drop-image RFC](../../implemented/simplification/2026-07-04-drop-image-content-block.md)). + - `isError: true` → map to the harness `isError` result path (`{ content: [...], isError: true }`). +3. Cancellation: `exec.signal` (from the agent loop's cancel) is passed through to the MCP SDK's `callTool`, which sends `$/cancelRequest` to the server. + +### Subprocess environment (stdio transport) + +Replicate the `buildChildEnv` + `SENSITIVE_ENV_PATTERN` scrub from `dsh-subagent-acp`: filter ambient env (strip credential-shaped vars matching `/KEY|SECRET|TOKEN/i`), then merge `config.env` on top. Explicit env overrides survive the scrub. + +### Disconnection / crash + +No auto-reconnect. If the MCP server process exits or the transport closes: + +1. The effect disposes → all registered tools are unregistered (fiber-scoped disposers). +2. Subsequent model calls to those tools → `ToolNotFoundError` → `isError: true`. +3. Recovery: user edits `cordis.yml` (triggers HMR reload) or restarts the harness. + +This matches the ACP subagent pattern: "crash = terminal, report error, clean up, don't retry." + +## Alternatives considered + +### MCP Server side (expose harness tools to external MCP clients) + +Deferred. The ACP bridge already exposes the harness as an agent server. Adding an MCP server layer would duplicate that with a different protocol, and the primary user need is consuming external tools, not exposing them. + +### Capability-seam three-package split (interface / impl / consumer) + +Rejected. There is no foreseeable alternative MCP client implementation — MCP has one protocol, one SDK. The convention is "don't split preemptively" until a second implementation appears. + +### Auto-reconnect with exponential backoff + +Rejected for v1. Adds complexity (partial-availability state where tools are registered but temporarily non-functional), and stdio process crashes usually indicate a configuration problem that retrying won't fix. HMR already provides the manual recovery path. Can be added as a future `reconnect: boolean` config if needed. + +### Bridge Resources and Prompts + +Deferred. Resources need a harness-side mechanism to decide WHEN to inject content (system prompt? on demand? model-triggered?). Prompts need a "prompt template" concept the harness lacks. Both require their own design; Tools are the high-value, low-risk starting point. + +### Raw model-facing tool names with an optional `toolPrefix` + +Rejected — this was the original proposal, built on the premise that "most MCP servers already use semantic prefixes in their tool names (e.g. `github_create_issue`)". The premise is false: the official GitHub server publishes `create_issue`, the reference filesystem server `read_file`, Sentry `search_issues` — and the Microsoft survey above shows collisions are common at ecosystem scale. Collision-time prefixing (or warn-and-skip) also makes the available tool set depend on plugin load order, and a tool could be silently renamed when an unrelated server is added — invalidating session history and permission rules mid-conversation. No surveyed multi-server agent product ships raw names. + +### Server-only namespace (`github__create_issue`, no `mcp__` marker) + +Rejected for v1. It prevents cross-server collisions but does not separate MCP registrations from native harness tools, and it forfeits MCP-wide policy shapes (`mcp__*`). The marker costs 5 characters; the `mcp____` spelling matches Claude Code and Codex, maximizing model familiarity. If the ToolRegistry later grows source-aware namespaces, dropping the literal marker can be revisited as a naming-policy change. + +### Deriving the namespace from the server-announced `serverInfo.name` + +Rejected. The remote name is untrusted, non-unique across deployments, and changeable on upgrade; tool identity and permission rules must not silently follow it. The namespace is local configuration. + +### Preserve multiple TextBlocks in tool result + +Rejected. `flattenText()` in the DeepSeek serializer uses `join('')` (no separator) when flattening `ContentBlock[]` to wire format. Multiple text blocks would silently lose inter-block boundaries — a correctness bug. All existing tools return a single TextBlock; the MCP bridge follows suit. + +## Testing + +Coverage is named per tier; each behavior lives at the cheapest tier that can express it. + +- **Unit** (`tests/mcp-client.spec.ts`, `tests/apply.spec.ts`, mocked MCP SDK): the `publicToolName` algorithm (clean, normalize, truncate-and-hash, determinism, distinct-identity separation), raw-vs-public wire discipline, cross-server and native-tool coexistence, duplicate-`serverName` load failure and reservation release, invalid-tool-list rejection, generation swap/rollback, failed-re-sync retention, result mapping, cancellation, config schema validation. 100% per-file coverage gates the package. +- **E2E** (`tests/mcp-client.e2e.ts`, keyless): the real MCP protocol against the in-repo fixture server, `@modelcontextprotocol/server-everything`, and `@modelcontextprotocol/server-filesystem` over stdio, and against an in-process `StreamableHTTPServerTransport` server over Streamable HTTP — discovery under the namespace, dotted-name normalization end to end, execution round-trips, duplicate-`serverName` rejection, disposal. +- **Snapshot**: deliberately none. MCP tools introduce no new transcript surface — they register as raw `ToolDefinition`s and render through the ACP bridge's generic-card fallback, which the bridge's unit suite already pins (`packages/ui/acp/tests/stream-update.spec.ts`). Adding an MCP server to the snapshot example's `cordis.yml` would mutate the pinned `text-turn` system-prompt fixture (forcing a with-key re-record of every recorded golden) and make every replay depend on spawning an external MCP server process — for zero new rendering behavior. If a later change gives MCP tools their own render intent, that change names its snapshot coverage then. + +## Consequences + +- A `cordis.yml` entry per MCP server is the entire integration cost: `serverName: filesystem` + a stdio command (or a Streamable HTTP URL) puts `mcp__filesystem__read_file` in the model's tool list, callable, with the raw `read_file` on the wire. +- Public names are part of session history and permission/config surfaces; the naming algorithm is a v1 contract pinned by tests, and changing it after release is a breaking change. +- The `mcp____` qualifier costs tokens on every name. Accepted: descriptions and JSON schemas dominate tool-definition tokens, and the qualifier buys stable identity, collision isolation, and MCP-wide policy shapes (`mcp__*`, `mcp__github__*`). +- **MCP SDK stability**: the `@modelcontextprotocol/sdk` is still evolving; breaking changes require updating the bridge. The version is pinned, and the SDK is widely adopted (Claude Desktop, Cursor, VS Code) so breaking changes are unlikely to be silent. +- **Tool schema quality**: MCP servers may expose poorly-described tools (vague descriptions, incomplete JSON schemas). The harness passes them through as-is — garbage-in-garbage-out; that is the server author's responsibility, not the bridge's. +- **Stdio process management**: a misbehaving MCP server that ignores signals could wedge dispose. The Cordis fiber disposal has bounded quiescence; a stuck transport eventually times out at the framework level. +- Crash recovery is manual (HMR edit or restart) — accepted for v1; a `reconnect` config remains open as future work. diff --git a/docs/rfc/proposed/feature/2026-07-07-mcp-client-plugin.md b/docs/rfc/proposed/feature/2026-07-07-mcp-client-plugin.md deleted file mode 100644 index 952335f840..0000000000 --- a/docs/rfc/proposed/feature/2026-07-07-mcp-client-plugin.md +++ /dev/null @@ -1,160 +0,0 @@ -# RFC: MCP client plugin — connect to external MCP servers and bridge their tools - -Status: proposed - -## Problem - -The harness has no way to consume tools from the MCP (Model Context Protocol) ecosystem. MCP is the emerging standard for tool servers — GitHub, filesystem, databases, code search, and hundreds of community servers expose tools via MCP. Users want to point the harness at one or more MCP servers and have their tools appear as native model-facing tools, without writing per-server glue code. - -The `ToolRegistry` already accepts raw JSON Schema tool definitions (documented in `dsh-tools` README: "Raw JSON-Schema tool definitions (from MCP servers) are still accepted by `ToolRegistry.register()` directly"), and the extension cookbook sketches the intended pattern ("MCP | one plugin per server: discover tools → `ctx.tools.register()`"). The infrastructure is ready; the bridge plugin is missing. - -## Proposal - -### Package - -A single package `@deepseek-ai/dsh-mcp-client` at `packages/mcp/mcp-client/`. No capability-seam three-package split — there is no foreseeable second MCP client implementation, and the convention is "don't split preemptively" ([capability seams RFC](../../implemented/architecture/2026-06-13-capability-seams.md)). - -### SDK - -Use the official [`@modelcontextprotocol/sdk`](https://github.com/modelcontextprotocol/typescript-sdk) (`Client`, `StdioClientTransport`, `StreamableHTTPClientTransport`). The harness does not implement its own JSON-RPC — consistent with how ACP delegates to `@agentclientprotocol/sdk`. - -### Scope - -MCP Client only (no server side — ACP already covers the "expose harness as an agent" role). Bridge **Tools** only — Resources and Prompts are deferred (they require harness-side consumption mechanisms that don't exist yet, and design space is large). - -### Plugin shape - -Namespace plugin (named exports `name`/`inject`/`Config`/`apply`, no `export default`). `inject: ['tools']`. Each MCP server is one plugin instance in `cordis.yml` — the same package loaded N times with different configs, like `dsh-tool-subagent`. - -### Configuration - -Flat discriminated union on the `transport` field: - -```typescript -interface StdioConfig { - transport: 'stdio' - command: string - args?: string[] - env?: Record - cwd?: string - toolPrefix?: string - toolCallTimeoutMs?: number // default 60_000 -} - -interface StreamableHttpConfig { - transport: 'streamable-http' - url: string - headers?: Record - toolPrefix?: string - toolCallTimeoutMs?: number // default 60_000 -} - -type Config = StdioConfig | StreamableHttpConfig -``` - -Example `cordis.yml` usage: - -```yaml -- id: mcp-github - name: '@deepseek-ai/dsh-mcp-client' - config: - transport: stdio - command: npx - args: ['-y', '@modelcontextprotocol/server-github'] - env: - GITHUB_TOKEN: !!js process.env.GITHUB_TOKEN - -- id: mcp-web - name: '@deepseek-ai/dsh-mcp-client' - config: - transport: streamable-http - url: http://localhost:3000/mcp - headers: - Authorization: !!js `Bearer ${process.env.MCP_TOKEN}` -``` - -### Lifecycle - -Boot-time from `cordis.yml`. HMR (`@cordisjs/plugin-hmr`) provides hot-swap: editing the yml entry triggers dispose of the old instance (disconnects, unregisters tools) and creation of a new one (connects, discovers, registers). No runtime-dynamic API for now. - -### Tool discovery and registration - -1. On connect: `client.listTools()` → register each tool as a raw `ToolDefinition` via `ctx.tools.register()`. -2. Listen for `notifications/tools/list_changed` → re-run `listTools()`, diff, unregister removed, register added. -3. Registration uses the raw JSON Schema from MCP (no `defineTool` DSL conversion). -4. No `presentCall`/`presentResult` — the ACP bridge's generic-card fallback handles rendering. -5. Tools are transparent in the system prompt — no "[via MCP]" annotation. - -### Name conflict handling - -If `config.toolPrefix` is set (e.g. `"gh_"`), it is prepended to each MCP tool name before registration. If a name collides with an already-registered tool, log a warning and skip that tool (do not crash the entire server connection). - -### Tool execution - -A unified `execute` handler for all tools from one MCP server: - -1. Call `client.callTool({ name, arguments }, { signal: exec.signal })` with the configured timeout. -2. Map the result: - - Multiple `text` content blocks → join with `'\n'` into a single `TextBlock` (required: `flattenText` uses `join('')` without separator, so multiple blocks would lose inter-block boundaries). - - `image` content blocks → discard with a `ctx.logger.warn` (the harness has no image content block type; [drop-image RFC](../../implemented/simplification/2026-07-04-drop-image-content-block.md)). - - `isError: true` → map to the harness `isError` result path (`{ content: [...], isError: true }`). -3. Cancellation: `exec.signal` (from the agent loop's cancel) is passed through to the MCP SDK's `callTool`, which sends `$/cancelRequest` to the server. - -### Subprocess environment (stdio transport) - -Replicate the `buildChildEnv` + `SENSITIVE_ENV_PATTERN` scrub from `dsh-subagent-acp`: filter ambient env (strip credential-shaped vars matching `/KEY|SECRET|TOKEN/i`), then merge `config.env` on top. Explicit env overrides survive the scrub. - -### Disconnection / crash - -No auto-reconnect. If the MCP server process exits or the transport closes: - -1. The effect disposes → all registered tools are unregistered (fiber-scoped disposers). -2. Subsequent model calls to those tools → `ToolNotFoundError` → `isError: true`. -3. Recovery: user edits `cordis.yml` (triggers HMR reload) or restarts the harness. - -This matches the ACP subagent pattern: "crash = terminal, report error, clean up, don't retry." - -## Alternatives considered - -### MCP Server side (expose harness tools to external MCP clients) - -Deferred. The ACP bridge already exposes the harness as an agent server. Adding an MCP server layer would duplicate that with a different protocol, and the primary user need is consuming external tools, not exposing them. - -### Capability-seam three-package split (interface / impl / consumer) - -Rejected. There is no foreseeable alternative MCP client implementation — MCP has one protocol, one SDK. The convention is "don't split preemptively" until a second implementation appears. - -### Auto-reconnect with exponential backoff - -Rejected for v1. Adds complexity (partial-availability state where tools are registered but temporarily non-functional), and stdio process crashes usually indicate a configuration problem that retrying won't fix. HMR already provides the manual recovery path. Can be added as a future `reconnect: boolean` config if needed. - -### Bridge Resources and Prompts - -Deferred. Resources need a harness-side mechanism to decide WHEN to inject content (system prompt? on demand? model-triggered?). Prompts need a "prompt template" concept the harness lacks. Both require their own design; Tools are the high-value, low-risk starting point. - -### Always-on namespace prefix (e.g. `mcp_github__create_issue`) - -Rejected. Most MCP servers already use semantic prefixes in their tool names (e.g. `github_create_issue`). A forced prefix would break model familiarity with well-known MCP tool names and waste context tokens. Optional `toolPrefix` handles the rare collision case. - -### Preserve multiple TextBlocks in tool result - -Rejected. `flattenText()` in the DeepSeek serializer uses `join('')` (no separator) when flattening `ContentBlock[]` to wire format. Multiple text blocks would silently lose inter-block boundaries — a correctness bug. All existing tools return a single TextBlock; the MCP bridge follows suit. - -## Acceptance criteria - -- A `cordis.yml` entry connecting to an MCP stdio server (e.g. `@modelcontextprotocol/server-filesystem`) results in that server's tools appearing in the model's tool list and being callable. -- A `cordis.yml` entry connecting via Streamable HTTP works equivalently. -- Adding/removing an MCP entry in `cordis.yml` while HMR is active hot-swaps the tools without restart. -- `toolPrefix` config correctly prefixes tool names; a name collision logs a warning and skips. -- Agent cancel propagates to in-flight `callTool` (abort signal). -- Timeout fires and produces an `isError` result when an MCP server hangs. -- Server crash cleanly unregisters tools (no orphaned tool definitions). -- `notifications/tools/list_changed` triggers a re-sync of tool registrations. -- 100% test coverage on the new package (unit tests with mocked MCP SDK). - -## Risks - -- **MCP SDK stability**: the `@modelcontextprotocol/sdk` is still evolving. Breaking changes in the SDK require updating the bridge. Mitigation: pin a specific version; the SDK is widely adopted (Claude Desktop, Cursor, VS Code) so breaking changes are unlikely to be silent. -- **Tool schema quality**: MCP servers may expose poorly-described tools (vague descriptions, incomplete JSON schemas). The harness passes them through as-is — garbage-in-garbage-out. Mitigation: this is the server author's responsibility, not the bridge's. -- **Stdio process management**: a misbehaving MCP server that ignores signals could wedge dispose. Mitigation: the Cordis fiber disposal has bounded quiescence; a stuck transport eventually times out at the framework level. -- **Token budget pressure**: connecting many MCP servers with many tools inflates the system prompt. Mitigation: no different from registering many native tools; the compaction layer handles context pressure. diff --git a/packages/mcp/mcp-client/README.md b/packages/mcp/mcp-client/README.md index 68f6ebf42a..ec7c615d04 100644 --- a/packages/mcp/mcp-client/README.md +++ b/packages/mcp/mcp-client/README.md @@ -1,6 +1,6 @@ # @deepseek-ai/dsh-mcp-client -MCP client bridge plugin: connects to external [Model Context Protocol](https://modelcontextprotocol.io/) servers and registers their tools on `ctx.tools`, making them available to the model as native tools. +MCP client bridge plugin: connects to external [Model Context Protocol](https://modelcontextprotocol.io/) servers and registers their tools on `ctx.tools`, making them available to the model as native tools under server-qualified names (`mcp____`). ## Usage @@ -10,6 +10,7 @@ One plugin instance per MCP server in `cordis.yml`: - id: mcp-github name: '@deepseek-ai/dsh-mcp-client' config: + serverName: github transport: stdio command: npx args: ['-y', '@modelcontextprotocol/server-github'] @@ -19,36 +20,45 @@ One plugin instance per MCP server in `cordis.yml`: - id: mcp-web name: '@deepseek-ai/dsh-mcp-client' config: + serverName: web transport: streamable-http url: http://localhost:3000/mcp headers: Authorization: !!js '`Bearer ${process.env.MCP_TOKEN}`' ``` -HMR hot-swaps: editing the entry triggers disconnect + reconnect without process restart. +The model sees `mcp__github__create_issue`, `mcp__web__search`, … — the same server-qualified shape Claude Code and Codex use. HMR hot-swaps: editing the entry triggers disconnect + reconnect without process restart; an unchanged `serverName` reproduces identical tool names. ## Config | Field | Transport | Required | Description | |---|---|---|---| | `transport` | both | yes | `"stdio"` or `"streamable-http"` | +| `serverName` | both | yes | Namespace for this server's model-facing tool names; `[A-Za-z0-9_-]{1,32}`, unique across live instances | | `command` | stdio | yes | Executable to spawn | | `args` | stdio | no | Arguments passed to the command | | `env` | stdio | no | Extra env vars merged on top of scrubbed ambient env | | `cwd` | stdio | no | Working directory for the child process | | `url` | http | yes | MCP server URL | | `headers` | http | no | Extra headers (e.g. auth tokens) | -| `toolPrefix` | both | no | Prefix prepended to each tool name before registration | | `toolCallTimeoutMs` | both | no | Timeout per `callTool` invocation (default 60000) | +## Tool naming + +Every MCP tool has two names: the raw MCP name (sent on the wire in `tools/call`) and the public name `mcp____` registered on `ctx.tools`. Public names are normalized to the DeepSeek function-name contract (64 chars, `[A-Za-z0-9_-]`); when replacement or truncation changes the name, a deterministic 12-hex-char hash of `(serverName, rawName)` is appended so distinct tools never collapse into one name. Names are pure functions of `(serverName, rawName)` — connection order, re-syncs, and other servers never rename a tool. + +- Two servers publishing the same raw name (e.g. `search`) coexist under their namespaces. +- A duplicate `serverName` across live instances fails the later plugin instance at load. +- A server listing the same tool name twice is rejected as an invalid tool list. +- A foreign registration squatting on this server's namespace rolls back the whole generation (never a partial set), with a loud error. + ## Behavior -- On connect: `listTools()` → registers each tool via `ctx.tools.register()`. -- Listens for `notifications/tools/list_changed` → re-syncs tool registrations. -- Tool execute: `client.callTool({ name, arguments }, { signal })` with timeout + abort support. -- Image content in results is discarded with a warning (the harness has no image block type). +- On connect: `listTools()` → registers each tool via `ctx.tools.register()` under its public name. +- Listens for `notifications/tools/list_changed` → re-syncs; a failed re-sync keeps the previous generation registered. +- Tool execute: `client.callTool({ name: rawName, arguments }, { signal })` with timeout + abort support — the public name is never sent to the server. +- Image content in results is discarded with a placeholder (the harness has no image block type). - On disconnect/crash: all tools are unregistered; no auto-reconnect. -- Name conflicts: if a tool name collides, it is skipped with a warning. Use `toolPrefix` to disambiguate. ## Services consumed diff --git a/packages/mcp/mcp-client/src/index.ts b/packages/mcp/mcp-client/src/index.ts index da3aefc3de..4a18f85ff5 100644 --- a/packages/mcp/mcp-client/src/index.ts +++ b/packages/mcp/mcp-client/src/index.ts @@ -1,11 +1,14 @@ /** * MCP client bridge plugin: connects to an external MCP server and registers - * its tools on `ctx.tools`. Each plugin instance connects to one MCP server; - * load multiple instances in `cordis.yml` for multiple servers. + * its tools on `ctx.tools` under server-qualified public names + * (`mcp____`). Each plugin instance connects to one MCP + * server; load multiple instances in `cordis.yml` for multiple servers. * * Namespace plugin (named exports, no default export). Lifecycle is - * effect-scoped: disposal disconnects from the server and unregisters all - * tools. HMR hot-swaps by disposing the old instance and creating a new one. + * effect-scoped: disposal disconnects from the server, unregisters all tools, + * and releases the `serverName` namespace reservation. HMR hot-swaps by + * disposing the old instance and creating a new one; identical `serverName` + * reproduces identical public tool names. * * @module @deepseek-ai/dsh-mcp-client */ @@ -28,12 +31,32 @@ export const inject = ['tools'] /** Default timeout for individual MCP tool calls (ms). */ const DEFAULT_TOOL_CALL_TIMEOUT_MS = 60_000 +/** + * Valid `serverName`: 1–32 chars of `[A-Za-z0-9_-]`. Kept well under the + * 64-char public-name budget so typical raw tool names survive unhashed. + */ +const SERVER_NAME_PATTERN = /^[A-Za-z0-9_-]{1,32}$/ + +/** + * Live `serverName` reservations per app, keyed off `ctx.root` (multiple apps + * in one process — tests — must not see each other's names). A duplicate + * namespace is a configuration error surfaced at plugin load, never silent + * shadowing. + */ +const activeServerNames = new WeakMap>() + // ---- Config ---- /** Config for connecting to an MCP server via a spawned child process over stdio. */ export interface StdioConfig { /** Transport type: spawn a child process and communicate over stdio. */ transport: 'stdio' + /** + * Stable local namespace for this server's model-facing tool names + * (`mcp____`). Must match `[A-Za-z0-9_-]{1,32}` and be + * unique across live mcp-client instances. + */ + serverName: string /** Executable to spawn. */ command: string /** Arguments passed to the command. */ @@ -42,8 +65,6 @@ export interface StdioConfig { env: Record /** Working directory for the child process. */ cwd: string - /** Prefix prepended to each tool name before registration. */ - toolPrefix: string /** Timeout per callTool invocation (ms). */ toolCallTimeoutMs: number } @@ -52,12 +73,16 @@ export interface StdioConfig { export interface StreamableHttpConfig { /** Transport type: connect to an MCP server over Streamable HTTP (SSE). */ transport: 'streamable-http' + /** + * Stable local namespace for this server's model-facing tool names + * (`mcp____`). Must match `[A-Za-z0-9_-]{1,32}` and be + * unique across live mcp-client instances. + */ + serverName: string /** MCP server URL. */ url: string /** Extra headers (e.g. auth tokens). */ headers: Record - /** Prefix prepended to each tool name before registration. */ - toolPrefix: string /** Timeout per callTool invocation (ms). */ toolCallTimeoutMs: number } @@ -68,18 +93,18 @@ export type Config = StdioConfig | StreamableHttpConfig export const Config = z.union([ z.object({ transport: z.const('stdio'), + serverName: z.string().required().pattern(SERVER_NAME_PATTERN), command: z.string().required(), args: z.array(String).default([]), env: z.dict(String).default({}), cwd: z.string().default(''), - toolPrefix: z.string().default(''), toolCallTimeoutMs: z.number().default(DEFAULT_TOOL_CALL_TIMEOUT_MS), }), z.object({ transport: z.const('streamable-http'), + serverName: z.string().required().pattern(SERVER_NAME_PATTERN), url: z.string().required(), headers: z.dict(String).default({}), - toolPrefix: z.string().default(''), toolCallTimeoutMs: z.number().default(DEFAULT_TOOL_CALL_TIMEOUT_MS), }), ]) as unknown as z @@ -87,42 +112,66 @@ export const Config = z.union([ // ---- Plugin apply ---- export function apply(ctx: Context, config: Config): void { + // Reserve the namespace first: a duplicate `serverName` fails THIS instance + // at load with an actionable error and leaves the earlier instance intact. + ctx.effect(() => { + let names = activeServerNames.get(ctx.root) + if (!names) { + names = new Set() + activeServerNames.set(ctx.root, names) + } + if (names.has(config.serverName)) { + throw new Error( + `mcp-client: serverName "${config.serverName}" is already in use by another mcp-client instance — pick a unique serverName in cordis.yml`, + ) + } + names.add(config.serverName) + return () => void names.delete(config.serverName) + }, 'mcp-client.serverName') + const transport = createTransport(config) const client = new Client( { name: 'dsh-mcp-client', version: '0.0.1' }, { capabilities: {} }, ) - // Connect and set up tools. Errors during connect are logged, not thrown - // (the plugin simply has no tools registered). + const opts = { + serverName: config.serverName, + toolCallTimeoutMs: config.toolCallTimeoutMs, + } + + // Connect and set up tools. Errors during connect/first sync are logged, + // not thrown (the plugin simply has no tools registered). `ready` resolves + // to an accessor for the CURRENT disposer generation, so the effect + // disposer below always unregisters the live set, not the first one. const ready = (async () => { await client.connect(transport) - let disposers = await syncTools(client, ctx, { - toolPrefix: config.toolPrefix, - toolCallTimeoutMs: config.toolCallTimeoutMs, - }, new Map()) + let disposers = await syncTools(client, ctx, opts, new Map()) client.setNotificationHandler( ToolListChangedNotificationSchema, async () => { - ctx.logger.info('mcp-client: tool list changed, re-syncing') - disposers = await syncTools(client, ctx, { - toolPrefix: config.toolPrefix, - toolCallTimeoutMs: config.toolCallTimeoutMs, - }, disposers) + ctx.logger.info(`mcp-client(${config.serverName}): tool list changed, re-syncing`) + try { + disposers = await syncTools(client, ctx, opts, disposers) + } catch (error) { + // Fetch-phase failure: the previous generation is still registered + // and `disposers` still owns it — keep serving the last good list. + ctx.logger.error(`mcp-client(${config.serverName}): tool re-sync failed: ${String(error)}`) + } }, ) - return disposers + return () => disposers })().catch((error: unknown) => { - ctx.logger.error(`mcp-client: failed to connect: ${String(error)}`) - return new Map void>() + ctx.logger.error(`mcp-client(${config.serverName}): failed to connect: ${String(error)}`) + return () => new Map void>() }) ctx.effect(() => async () => { - const disposers = await ready - for (const dispose of disposers.values()) dispose() + const live = await ready + for (const dispose of live().values()) dispose() try { await client.close() } catch { /* transport already gone */ } }, 'mcp-client.connection') } diff --git a/packages/mcp/mcp-client/src/tools.ts b/packages/mcp/mcp-client/src/tools.ts index c3a35ccfba..ae01fc0f84 100644 --- a/packages/mcp/mcp-client/src/tools.ts +++ b/packages/mcp/mcp-client/src/tools.ts @@ -1,37 +1,87 @@ /** - * Tool bridge: discovers MCP tools, registers them on the harness ToolRegistry, - * and handles re-sync when the server's tool list changes. + * Tool bridge: discovers MCP tools, registers them on the harness ToolRegistry + * under deterministic server-qualified public names, and handles re-sync when + * the server's tool list changes. + * + * Naming contract (see the mcp-client RFC "Naming invariants"): every MCP tool + * has the stable identity `(serverName, rawName)`; the model-facing public name + * is `mcp____`, normalized to the DeepSeek function-name + * constraints. The raw name is only ever sent on the wire (`tools/call`); the + * public name is never parsed to recover it. * * @module */ +import { createHash } from 'node:crypto' import type { Client } from '@modelcontextprotocol/sdk/client/index.js' import type { Context } from 'cordis' import type { ToolDefinition, ToolExecution } from '@deepseek-ai/dsh-tools' /** Resolved options relevant to tool bridging. */ export interface ToolBridgeOptions { - toolPrefix: string + serverName: string toolCallTimeoutMs: number } -/** State for one sync generation: the current set of disposers keyed by tool name. */ -type ToolDisposers = Map void> +/** State for one sync generation: the current set of disposers keyed by public name. */ +export type ToolDisposers = Map void> + +/** + * DeepSeek function-name contract: at most 64 characters. Wire-protocol + * constant, not configuration. + */ +const MAX_PUBLIC_NAME_LENGTH = 64 + +/** DeepSeek function-name contract: only `[A-Za-z0-9_-]` is allowed. */ +const INVALID_NAME_CHARS = /[^A-Za-z0-9_-]/g + +/** Hex chars of the SHA-256 identity hash appended on lossy normalization. */ +const HASH_LENGTH = 12 + +/** + * Derive the model-facing public name for one MCP tool. + * + * Deterministic pure function of `(serverName, rawName)`: the clean case is + * `mcp____` verbatim. When character replacement or + * truncation to the DeepSeek function-name contract (64 chars, + * `[A-Za-z0-9_-]`) changes the name, a 12-hex-char SHA-256 hash of the + * identity is appended so distinct MCP identities never collapse into the + * same public name. + * + * @param serverName - Stable local namespace from plugin config. + * @param rawName - The MCP server's own tool name. + * @returns The globally unique, model-facing ToolRegistry name. + */ +export function publicToolName(serverName: string, rawName: string): string { + const joined = `mcp__${serverName}__${rawName}` + const normalized = joined.replace(INVALID_NAME_CHARS, '_') + if (normalized === joined && normalized.length <= MAX_PUBLIC_NAME_LENGTH) return normalized + const hash = createHash('sha256').update(`${serverName}\0${rawName}`).digest('hex').slice(0, HASH_LENGTH) + return `${normalized.slice(0, MAX_PUBLIC_NAME_LENGTH - HASH_LENGTH - 1)}_${hash}` +} /** * Sync the MCP server's tool list into the harness ToolRegistry. * - * - Calls `client.listTools()` (paginated: drains all pages). - * - Registers each tool as a raw `ToolDefinition`. - * - On name conflict: logs a warning and skips that tool. - * - Returns a disposer map; call each value to unregister. + * Two phases keep the swap safe: + * + * 1. Fetch: drain `client.listTools()` pagination and build the full next + * generation of `ToolDefinition`s under public names. Any failure here + * (network error, duplicate raw name in the server's list) rejects and + * leaves the previous generation registered untouched. + * 2. Swap: dispose the previous generation, register the new one. A registry + * conflict here can only mean a foreign registration squats on this + * server's `mcp____` namespace — the partial generation is + * rolled back (zero tools from this server), the error is logged, and an + * empty map is returned. * * @param client - Connected MCP Client instance used to list and call tools. * @param ctx - Cordis context providing the `tools` service for registration. - * @param opts - Bridge options: tool name prefix and per-call timeout. - * @param previous - Disposer map from a prior sync generation; all entries are - * disposed before re-registering. - * @returns A map of registered tool names to their unregister disposers. + * @param opts - Bridge options: server namespace and per-call timeout. + * @param previous - Disposer map from the prior sync generation; disposed + * during the swap phase (only after the fetch phase succeeded). + * @returns A map of registered public tool names to their unregister + * disposers — the exact set of live registrations owned by this server. */ export async function syncTools( client: Client, @@ -39,32 +89,43 @@ export async function syncTools( opts: ToolBridgeOptions, previous: ToolDisposers, ): Promise { - for (const dispose of previous.values()) dispose() - - const disposers: ToolDisposers = new Map() - + // Phase 1: fetch and build the next generation without touching the registry. + const definitions = new Map() let cursor: string | undefined do { const response = await client.listTools(cursor ? { cursor } : undefined) for (const tool of response.tools) { - const registeredName = opts.toolPrefix + tool.name - const definition: ToolDefinition = { - name: registeredName, + const publicName = publicToolName(opts.serverName, tool.name) + if (definitions.has(publicName)) { + throw new Error( + `mcp-client(${opts.serverName}): server listed tool "${tool.name}" more than once — invalid tool list`, + ) + } + definitions.set(publicName, { + name: publicName, description: tool.description ?? '', parameters: tool.inputSchema, execute: createExecutor(client, tool.name, opts), - } - try { - const dispose = ctx.tools.register(definition) - disposers.set(registeredName, dispose) - } catch { - // Name conflict — another tool with this name is already registered. - ctx.logger.warn(`mcp-client: skipping tool "${registeredName}" (name conflict)`) - } + }) } cursor = response.nextCursor } while (cursor) + // Phase 2: swap generations. + for (const dispose of previous.values()) dispose() + const disposers: ToolDisposers = new Map() + try { + for (const [publicName, definition] of definitions) { + disposers.set(publicName, ctx.tools.register(definition)) + } + } catch (error) { + // A conflict on an `mcp____`-qualified name means a foreign + // registration occupies this server's namespace. Roll back so the model + // sees either the full generation or none of it — never a partial set. + for (const dispose of disposers.values()) dispose() + ctx.logger.error(`mcp-client(${opts.serverName}): tool registration failed, no tools registered: ${String(error)}`) + return new Map() + } return disposers } @@ -81,16 +142,17 @@ interface McpContentBlock { } /** - * Create an execute function for one MCP tool. The executor calls - * `client.callTool` with abort signal and timeout, then maps the result - * to harness ContentBlocks. + * Create an execute function for one MCP tool. The executor closes over the + * raw MCP tool name and calls `client.callTool` with it (never the public + * name), with abort signal and timeout, then maps the result to harness + * ContentBlocks. * * When the MCP server returns `isError: true`, the executor throws so that * the ToolRegistry's catch path produces an `isError` result for the model. */ function createExecutor( client: Client, - mcpToolName: string, + rawName: string, opts: ToolBridgeOptions, ): ToolDefinition['execute'] { return async (args: unknown, exec: ToolExecution) => { @@ -100,7 +162,7 @@ function createExecutor( // specific "missing required param" error the model can learn from. const argsObj = (typeof args === 'object' && args !== null ? args : {}) as Record const result = await client.callTool( - { name: mcpToolName, arguments: argsObj }, + { name: rawName, arguments: argsObj }, undefined, { ...exec.signal ? { signal: exec.signal } : {}, @@ -122,7 +184,7 @@ function createExecutor( // with optional fallbacks). // eslint-disable-next-line @typescript-eslint/no-unsafe-assignment const content: McpContentBlock[] = result.content - const text = extractText(content, mcpToolName) + const text = extractText(content, rawName) // MCP isError → throw so ToolRegistry produces an isError result for the model. if ('isError' in result && result.isError === true) { diff --git a/packages/mcp/mcp-client/tests/apply.spec.ts b/packages/mcp/mcp-client/tests/apply.spec.ts index 077cb4e3f6..4b43411346 100644 --- a/packages/mcp/mcp-client/tests/apply.spec.ts +++ b/packages/mcp/mcp-client/tests/apply.spec.ts @@ -10,19 +10,23 @@ import type { Config } from '@deepseek-ai/dsh-mcp-client' // ---- Mock MCP SDK ---- -const mockConnect = vi.fn<() => Promise>() -const mockClose = vi.fn<() => Promise>() -const mockListTools = vi.fn() -const mockCallTool = vi.fn() -const mockSetNotificationHandler = vi.fn() - -class MockClient { - connect = mockConnect - close = mockClose - listTools = mockListTools - callTool = mockCallTool - setNotificationHandler = mockSetNotificationHandler -} +// vi.mock factories are hoisted above every import/const, so the mock fns and +// class must be created inside vi.hoisted to exist when the factories run. +const { mockConnect, mockClose, mockListTools, mockCallTool, mockSetNotificationHandler, MockClient } = vi.hoisted(() => { + const mockConnect = vi.fn<() => Promise>() + const mockClose = vi.fn<() => Promise>() + const mockListTools = vi.fn() + const mockCallTool = vi.fn() + const mockSetNotificationHandler = vi.fn() + class MockClient { + connect = mockConnect + close = mockClose + listTools = mockListTools + callTool = mockCallTool + setNotificationHandler = mockSetNotificationHandler + } + return { mockConnect, mockClose, mockListTools, mockCallTool, mockSetNotificationHandler, MockClient } +}) vi.mock('@modelcontextprotocol/sdk/client/index.js', () => ({ Client: MockClient, @@ -36,11 +40,9 @@ vi.mock('@modelcontextprotocol/sdk/client/streamableHttp.js', () => ({ StreamableHTTPClientTransport: vi.fn(), })) -// ---- Import under test (after mocks) ---- - -const { apply, name, inject, Config: ConfigSchema } = await import( - '@deepseek-ai/dsh-mcp-client/src/index.ts', -) +// vi.mock is hoisted above static imports, so the module under test sees the +// mocked SDK even through a static import. +import { apply, name, inject, Config as ConfigSchema } from '@deepseek-ai/dsh-mcp-client/src/index.ts' // ---- Helpers ---- @@ -51,13 +53,22 @@ async function mountRegistry(): Promise { return ctx } +function sleep(ms: number): Promise { + // Annotated binding (not withResolvers()): the tests lint layer runs + // no-invalid-void-type with default options, which rejects the explicit + // type argument in call position but accepts the inferred form. + const gate: PromiseWithResolvers = Promise.withResolvers() + setTimeout(gate.resolve, ms) + return gate.promise +} + const stdioConfig: Config = { transport: 'stdio', + serverName: 'srv', command: 'echo', args: [], env: {}, cwd: '', - toolPrefix: '', toolCallTimeoutMs: 60_000, } @@ -69,6 +80,37 @@ describe('mcp-client plugin module exports', () => { expect(inject).toEqual(['tools']) expect(ConfigSchema).toBeDefined() }) + + it('Config schema rejects a missing serverName', () => { + expect(() => ConfigSchema({ + transport: 'stdio', + command: 'echo', + } as never)).toThrow() + }) + + it('Config schema rejects an invalid serverName', () => { + // schemastery unions wrap branch errors in a generic "expected ... but got" + // message, so assert the throw, not the inner pattern text. + expect(() => ConfigSchema({ + transport: 'stdio', + serverName: 'bad name!', + command: 'echo', + } as never)).toThrow() + expect(() => ConfigSchema({ + transport: 'stdio', + serverName: 'x'.repeat(33), + command: 'echo', + } as never)).toThrow() + }) + + it('Config schema accepts a valid serverName', () => { + const resolved = ConfigSchema({ + transport: 'stdio', + serverName: 'github-prod_1', + command: 'echo', + } as never) + expect(resolved.serverName).toBe('github-prod_1') + }) }) describe('apply (plugin lifecycle)', () => { @@ -86,39 +128,78 @@ describe('apply (plugin lifecycle)', () => { ctx = await mountRegistry() }) - it('connects, syncs tools, and registers a notification handler', async () => { + it('connects, syncs tools under the namespace, and registers a notification handler', async () => { apply(ctx, stdioConfig) - await new Promise(r => setTimeout(r, 50)) + await sleep(50) expect(mockConnect).toHaveBeenCalled() expect(mockListTools).toHaveBeenCalled() expect(mockSetNotificationHandler).toHaveBeenCalled() - expect(ctx.tools.get('remote')).toBeDefined() - }) - - it('applies toolPrefix from config during sync', async () => { - apply(ctx, { ...stdioConfig, toolPrefix: 'mcp_' }) - await new Promise(r => setTimeout(r, 50)) - - expect(ctx.tools.get('mcp_remote')).toBeDefined() + expect(ctx.tools.get('mcp__srv__remote')).toBeDefined() expect(ctx.tools.get('remote')).toBeUndefined() }) - it('logs error and registers no tools when connect fails', async () => { + it('rejects a duplicate serverName at load and leaves the first instance intact', async () => { + apply(ctx, stdioConfig) + await sleep(50) + expect(ctx.tools.get('mcp__srv__remote')).toBeDefined() + + expect(() => { apply(ctx, stdioConfig) }).toThrow(/serverName "srv" is already in use/) + // First instance unaffected. + expect(ctx.tools.get('mcp__srv__remote')).toBeDefined() + }) + + it('releases the serverName reservation on dispose', async () => { + const first = new Context() + await first.plugin(SystemPrompt) + await first.plugin(ToolRegistry) + apply(first, stdioConfig) + await sleep(50) + + await first.fiber.dispose() + await sleep(50) + + // Same root would conflict; a fresh app root reuses the name freely, + // and the disposed instance no longer holds the reservation on its root. + const second = new Context() + await second.plugin(SystemPrompt) + await second.plugin(ToolRegistry) + expect(() => { apply(second, stdioConfig) }).not.toThrow() + }) + + it('scopes serverName reservations per app root', async () => { + const other = await mountRegistry() + + apply(ctx, stdioConfig) + // Same serverName on a DIFFERENT root is fine. + expect(() => { apply(other, stdioConfig) }).not.toThrow() + await sleep(50) + + expect(ctx.tools.get('mcp__srv__remote')).toBeDefined() + expect(other.tools.get('mcp__srv__remote')).toBeDefined() + }) + + it('logs error and registers no tools when connect fails; dispose is a no-op', async () => { mockConnect.mockRejectedValue(new Error('connection refused')) apply(ctx, stdioConfig) - await new Promise(r => setTimeout(r, 50)) + await sleep(50) expect(mockListTools).not.toHaveBeenCalled() - expect(ctx.tools.get('remote')).toBeUndefined() + expect(ctx.tools.get('mcp__srv__remote')).toBeUndefined() + + // Disposal exercises the empty fallback accessor: nothing to unregister, + // close still attempted, no throw. + await ctx.fiber.dispose() + await sleep(50) + expect(mockClose).toHaveBeenCalled() }) it('re-syncs tools on ToolListChanged notification', async () => { apply(ctx, stdioConfig) - await new Promise(r => setTimeout(r, 50)) + await sleep(50) - expect(ctx.tools.get('remote')).toBeDefined() + expect(ctx.tools.get('mcp__srv__remote')).toBeDefined() // Simulate the notification handler being invoked with a new tool list. mockListTools.mockResolvedValue({ @@ -130,33 +211,55 @@ describe('apply (plugin lifecycle)', () => { const handler = mockSetNotificationHandler.mock.calls[0]![1] as () => Promise await handler() - expect(ctx.tools.get('remote')).toBeUndefined() - expect(ctx.tools.get('updated')).toBeDefined() + expect(ctx.tools.get('mcp__srv__remote')).toBeUndefined() + expect(ctx.tools.get('mcp__srv__updated')).toBeDefined() }) - it('effect disposer unregisters tools and closes client', async () => { + it('keeps the previous generation when a re-sync fails', async () => { apply(ctx, stdioConfig) - await new Promise(r => setTimeout(r, 50)) + await sleep(50) + expect(ctx.tools.get('mcp__srv__remote')).toBeDefined() - expect(ctx.tools.get('remote')).toBeDefined() + mockListTools.mockRejectedValue(new Error('flaky server')) + const handler = mockSetNotificationHandler.mock.calls[0]![1] as () => Promise + // Must not reject (contained), and must keep the last good generation. + await handler() - // Trigger disposal by disposing a child scope. - // Cordis ctx.effect registers the disposer; calling scope dispose runs it. - await ctx.fiber.dispose() - await new Promise(r => setTimeout(r, 50)) + expect(ctx.tools.get('mcp__srv__remote')).toBeDefined() + }) + + it('effect disposer unregisters the CURRENT generation and closes client', async () => { + // Load through ctx.plugin so ONLY the plugin's fiber is disposed — the + // registry must survive to observe the unregistration. + const fiber = ctx.plugin({ name: 'mcp-client', inject: ['tools'], apply }, stdioConfig) + await sleep(50) + + // Advance to a second generation first. + mockListTools.mockResolvedValue({ + tools: [{ name: 'updated', inputSchema: { type: 'object' } }], + nextCursor: undefined, + }) + const handler = mockSetNotificationHandler.mock.calls[0]![1] as () => Promise + await handler() + expect(ctx.tools.get('mcp__srv__updated')).toBeDefined() + + await fiber.dispose() + await sleep(50) expect(mockClose).toHaveBeenCalled() + // The live (second) generation was unregistered, not just the first. + expect(ctx.tools.get('mcp__srv__updated')).toBeUndefined() }) it('effect disposer handles client.close failure gracefully', async () => { mockClose.mockRejectedValue(new Error('already closed')) apply(ctx, stdioConfig) - await new Promise(r => setTimeout(r, 50)) + await sleep(50) // Should not throw when dispose is triggered. await ctx.fiber.dispose() - await new Promise(r => setTimeout(r, 50)) + await sleep(50) expect(mockClose).toHaveBeenCalled() }) @@ -164,16 +267,16 @@ describe('apply (plugin lifecycle)', () => { it('uses streamable-http config path', async () => { const httpConfig: Config = { transport: 'streamable-http', + serverName: 'web', url: 'http://localhost:3000/mcp', headers: { Authorization: 'Bearer x' }, - toolPrefix: '', toolCallTimeoutMs: 30_000, } apply(ctx, httpConfig) - await new Promise(r => setTimeout(r, 50)) + await sleep(50) expect(mockConnect).toHaveBeenCalled() - expect(ctx.tools.get('remote')).toBeDefined() + expect(ctx.tools.get('mcp__web__remote')).toBeDefined() }) }) diff --git a/packages/mcp/mcp-client/tests/fixture-server.ts b/packages/mcp/mcp-client/tests/fixture-server.ts index 6b97c2b59c..d127412736 100644 --- a/packages/mcp/mcp-client/tests/fixture-server.ts +++ b/packages/mcp/mcp-client/tests/fixture-server.ts @@ -51,5 +51,15 @@ server.registerTool('image', { ], })) +// Dotted name: legal in MCP, illegal in the DeepSeek function-name contract. +// Exercises the bridge's normalize-and-hash public-name path end to end. +server.registerTool('admin.reset', { + title: 'Admin Reset Tool', + description: 'Tool with a dotted name (normalization test).', + inputSchema: {}, +}, async () => ({ + content: [{ type: 'text', text: 'reset done' }], +})) + const transport = new StdioServerTransport() await server.connect(transport) diff --git a/packages/mcp/mcp-client/tests/mcp-client.e2e.ts b/packages/mcp/mcp-client/tests/mcp-client.e2e.ts index d59ba59d66..686d51acea 100644 --- a/packages/mcp/mcp-client/tests/mcp-client.e2e.ts +++ b/packages/mcp/mcp-client/tests/mcp-client.e2e.ts @@ -1,22 +1,29 @@ /** - * End-to-end tests for dsh-mcp-client. Exercises the REAL MCP protocol over - * stdio transport against: - * 1. A self-written fixture server (controlled edge cases) + * End-to-end tests for dsh-mcp-client. Exercises the REAL MCP protocol against: + * 1. A self-written fixture server over stdio (controlled edge cases) * 2. @modelcontextprotocol/server-everything (official integration test server) * 3. @modelcontextprotocol/server-filesystem (real filesystem operations) + * 4. An in-process StreamableHTTPServerTransport server over Streamable HTTP * * No API key needed — all servers are local/keyless. */ +import { createServer, type IncomingMessage, type Server, type ServerResponse } from 'node:http' import { mkdtemp, rm, writeFile, readFile } from 'node:fs/promises' import { tmpdir } from 'node:os' import { join } from 'node:path' import { fileURLToPath } from 'node:url' import { afterAll, beforeAll, describe, expect, it } from 'vitest' import { Context } from 'cordis' +import { McpServer } from '@modelcontextprotocol/sdk/server/mcp.js' +import { StreamableHTTPServerTransport } from '@modelcontextprotocol/sdk/server/streamableHttp.js' +import { z } from 'zod' +import type { Transport } from '@modelcontextprotocol/sdk/shared/transport.js' import SystemPrompt from '@deepseek-ai/dsh-system-prompt' import ToolRegistry from '@deepseek-ai/dsh-tools' import { CallId } from '@deepseek-ai/dsh-llm' +import { apply } from '@deepseek-ai/dsh-mcp-client/src/index.ts' +import { publicToolName } from '@deepseek-ai/dsh-mcp-client/src/tools.ts' import type { Config } from '@deepseek-ai/dsh-mcp-client' const tsxLoader = fileURLToPath(import.meta.resolve('tsx')) @@ -38,16 +45,31 @@ async function mountRegistry(): Promise { /** Apply the MCP client plugin and wait for tools to be registered. */ async function applyAndWait(ctx: Context, config: Config, timeoutMs = 20_000): Promise { - const { apply } = await import('@deepseek-ai/dsh-mcp-client/src/index.ts') - const toolsReady = new Promise((resolve, reject) => { - const timer = setTimeout( - () => { reject(new Error(`applyAndWait timed out after ${timeoutMs}ms — no tools/change event`)) }, - timeoutMs, - ) - ctx.on('tools/change', () => { clearTimeout(timer); resolve() }) - }) + // Annotated bindings (not withResolvers()): the tests lint layer runs + // no-invalid-void-type with default options, which rejects the explicit + // type argument in call position but accepts the inferred form. + const gate: PromiseWithResolvers = Promise.withResolvers() + const timer = setTimeout( + () => { gate.reject(new Error(`applyAndWait timed out after ${timeoutMs}ms — no tools/change event`)) }, + timeoutMs, + ) + ctx.on('tools/change', () => { clearTimeout(timer); gate.resolve() }) apply(ctx, config) - await toolsReady + await gate.promise +} + +function sleep(ms: number): Promise { + const gate: PromiseWithResolvers = Promise.withResolvers() + setTimeout(gate.resolve, ms) + return gate.promise +} + +/** Narrow a result content block to its text, failing the test on any other shape. */ +function textOf(block: unknown): string { + if (block && typeof block === 'object' && 'text' in block && typeof block.text === 'string') { + return block.text + } + throw new Error(`expected a text content block, got ${JSON.stringify(block)}`) } let callSeq = 0 @@ -62,11 +84,11 @@ describe('fixture server — controlled scenarios', () => { const fixtureConfig: Config = { transport: 'stdio', + serverName: 'fixture', command: process.execPath, args: ['--import', tsxLoader, fixtureServerPath], env: { TSX_TSCONFIG_PATH: repoTsconfig }, cwd: packageDir, - toolPrefix: '', toolCallTimeoutMs: 15_000, } @@ -77,21 +99,37 @@ describe('fixture server — controlled scenarios', () => { afterAll(async () => { if (ctx) await ctx.fiber.dispose() - await new Promise(r => setTimeout(r, 200)) + await sleep(200) }) - it('discovers all fixture tools', () => { + it('discovers all fixture tools under the server namespace', () => { const schemas = ctx.tools.schemas() const names = schemas.map(s => s.name) - expect(names).toContain('add') - expect(names).toContain('greet') - expect(names).toContain('fail') - expect(names).toContain('image') + expect(names).toContain('mcp__fixture__add') + expect(names).toContain('mcp__fixture__greet') + expect(names).toContain('mcp__fixture__fail') + expect(names).toContain('mcp__fixture__image') + // Raw names are not registered. + expect(names).not.toContain('add') + }) + + it('normalizes the dotted tool name with a deterministic hash suffix', () => { + const publicName = publicToolName('fixture', 'admin.reset') + expect(publicName).toMatch(/^mcp__fixture__admin_reset_[0-9a-f]{12}$/) + expect(ctx.tools.get(publicName)).toBeDefined() + }) + + it('executes the dotted tool via its normalized public name', async () => { + const result = await ctx.tools.execute({ + callId: nextCallId(), name: publicToolName('fixture', 'admin.reset'), arguments: {}, + }) + expect(result.isError).toBe(false) + expect(result.content[0]).toEqual({ type: 'text', text: 'reset done' }) }) it('executes add(2, 3) → "5"', async () => { const result = await ctx.tools.execute({ - callId: nextCallId(), name: 'add', arguments: { a: 2, b: 3 }, + callId: nextCallId(), name: 'mcp__fixture__add', arguments: { a: 2, b: 3 }, }) expect(result.isError).toBe(false) expect(result.content[0]).toEqual({ type: 'text', text: '5' }) @@ -99,7 +137,7 @@ describe('fixture server — controlled scenarios', () => { it('executes greet("World") → "Hello, World!"', async () => { const result = await ctx.tools.execute({ - callId: nextCallId(), name: 'greet', arguments: { name: 'World' }, + callId: nextCallId(), name: 'mcp__fixture__greet', arguments: { name: 'World' }, }) expect(result.isError).toBe(false) expect(result.content[0]).toEqual({ type: 'text', text: 'Hello, World!' }) @@ -107,7 +145,7 @@ describe('fixture server — controlled scenarios', () => { it('executes fail() → isError result', async () => { const result = await ctx.tools.execute({ - callId: nextCallId(), name: 'fail', arguments: {}, + callId: nextCallId(), name: 'mcp__fixture__fail', arguments: {}, }) expect(result.isError).toBe(true) expect(result.content[0]).toMatchObject({ type: 'text' }) @@ -115,49 +153,35 @@ describe('fixture server — controlled scenarios', () => { it('executes image() → image placeholder', async () => { const result = await ctx.tools.execute({ - callId: nextCallId(), name: 'image', arguments: {}, + callId: nextCallId(), name: 'mcp__fixture__image', arguments: {}, }) expect(result.isError).toBe(false) - const text = (result.content[0] as { type: string; text: string }).text + const text = textOf(result.content[0]) expect(text).toContain('Here is an image:') expect(text).toContain('[image: image/png, content discarded]') expect(text).toContain('End of image.') }) }) -describe('fixture server — toolPrefix', () => { - let ctx: Context - - beforeAll(async () => { - ctx = await mountRegistry() - await applyAndWait(ctx, { +describe('fixture server — duplicate serverName', () => { + it('rejects a second instance with the same serverName on one root', async () => { + const ctx = await mountRegistry() + const config: Config = { transport: 'stdio', + serverName: 'dup', command: process.execPath, args: ['--import', tsxLoader, fixtureServerPath], env: { TSX_TSCONFIG_PATH: repoTsconfig }, cwd: packageDir, - toolPrefix: 'fx_', toolCallTimeoutMs: 15_000, - }) + } + await applyAndWait(ctx, config) + + expect(() => { apply(ctx, config) }).toThrow(/serverName "dup" is already in use/) + + await ctx.fiber.dispose() + await sleep(200) }, 30_000) - - afterAll(async () => { - if (ctx) await ctx.fiber.dispose() - await new Promise(r => setTimeout(r, 200)) - }) - - it('registers tools with prefix', () => { - expect(ctx.tools.get('fx_add')).toBeDefined() - expect(ctx.tools.get('fx_greet')).toBeDefined() - expect(ctx.tools.get('add')).toBeUndefined() - }) - - it('executes prefixed tool', async () => { - const result = await ctx.tools.execute({ - callId: nextCallId(), name: 'fx_add', arguments: { a: 10, b: 20 }, - }) - expect(result.content[0]).toEqual({ type: 'text', text: '30' }) - }) }) describe('fixture server — disposal', () => { @@ -165,21 +189,21 @@ describe('fixture server — disposal', () => { const ctx = await mountRegistry() await applyAndWait(ctx, { transport: 'stdio', + serverName: 'fixture', command: process.execPath, args: ['--import', tsxLoader, fixtureServerPath], env: { TSX_TSCONFIG_PATH: repoTsconfig }, cwd: packageDir, - toolPrefix: '', toolCallTimeoutMs: 15_000, }) // Tools are registered before dispose. - expect(ctx.tools.get('add')).toBeDefined() + expect(ctx.tools.get('mcp__fixture__add')).toBeDefined() expect(ctx.tools.schemas().length).toBeGreaterThanOrEqual(4) // Dispose should complete without throwing. await ctx.fiber.dispose() - await new Promise(r => setTimeout(r, 200)) + await sleep(200) }, 30_000) }) @@ -190,11 +214,11 @@ describe('server-everything — official test server', () => { const config: Config = { transport: 'stdio', + serverName: 'everything', command: join(localBin, 'mcp-server-everything'), args: ['stdio'], env: {}, cwd: '', - toolPrefix: '', toolCallTimeoutMs: 30_000, } @@ -205,43 +229,40 @@ describe('server-everything — official test server', () => { afterAll(async () => { if (ctx) await ctx.fiber.dispose() - await new Promise(r => setTimeout(r, 500)) + await sleep(500) }) it('discovers tools from server-everything', () => { const schemas = ctx.tools.schemas() const names = schemas.map(s => s.name) - expect(names).toContain('echo') - expect(names).toContain('get-sum') - expect(names).toContain('get-tiny-image') + expect(names).toContain('mcp__everything__echo') + expect(names).toContain('mcp__everything__get-sum') + expect(names).toContain('mcp__everything__get-tiny-image') expect(names.length).toBeGreaterThanOrEqual(8) }) it('executes echo({ message: "hello" }) → "Echo: hello"', async () => { const result = await ctx.tools.execute({ - callId: nextCallId(), name: 'echo', arguments: { message: 'hello' }, + callId: nextCallId(), name: 'mcp__everything__echo', arguments: { message: 'hello' }, }) expect(result.isError).toBe(false) - const text = (result.content[0] as { type: string; text: string }).text - expect(text).toBe('Echo: hello') + expect(textOf(result.content[0])).toBe('Echo: hello') }) it('executes get-sum({ a: 3, b: 7 }) → contains "10"', async () => { const result = await ctx.tools.execute({ - callId: nextCallId(), name: 'get-sum', arguments: { a: 3, b: 7 }, + callId: nextCallId(), name: 'mcp__everything__get-sum', arguments: { a: 3, b: 7 }, }) expect(result.isError).toBe(false) - const text = (result.content[0] as { type: string; text: string }).text - expect(text).toContain('10') + expect(textOf(result.content[0])).toContain('10') }) it('executes get-tiny-image → image placeholder', async () => { const result = await ctx.tools.execute({ - callId: nextCallId(), name: 'get-tiny-image', arguments: {}, + callId: nextCallId(), name: 'mcp__everything__get-tiny-image', arguments: {}, }) expect(result.isError).toBe(false) - const text = (result.content[0] as { type: string; text: string }).text - expect(text).toContain('[image: image/png, content discarded]') + expect(textOf(result.content[0])).toContain('[image: image/png, content discarded]') }) }) @@ -257,11 +278,11 @@ describe('server-filesystem — real filesystem operations', () => { ctx = await mountRegistry() const config: Config = { transport: 'stdio', + serverName: 'filesystem', command: join(localBin, 'mcp-server-filesystem'), args: [tempDir], env: {}, cwd: '', - toolPrefix: '', toolCallTimeoutMs: 30_000, } await applyAndWait(ctx, config) @@ -269,16 +290,16 @@ describe('server-filesystem — real filesystem operations', () => { afterAll(async () => { if (ctx) await ctx.fiber.dispose() - await new Promise(r => setTimeout(r, 500)) + await sleep(500) await rm(tempDir, { recursive: true, force: true }) }) it('discovers filesystem tools', () => { const schemas = ctx.tools.schemas() const names = schemas.map(s => s.name) - expect(names).toContain('read_file') - expect(names).toContain('write_file') - expect(names).toContain('list_directory') + expect(names).toContain('mcp__filesystem__read_file') + expect(names).toContain('mcp__filesystem__write_file') + expect(names).toContain('mcp__filesystem__list_directory') }) it('write_file + read_file round-trip', async () => { @@ -287,7 +308,7 @@ describe('server-filesystem — real filesystem operations', () => { // Write via MCP tool const writeResult = await ctx.tools.execute({ - callId: nextCallId(), name: 'write_file', arguments: { path: filePath, content }, + callId: nextCallId(), name: 'mcp__filesystem__write_file', arguments: { path: filePath, content }, }) expect(writeResult.isError).toBe(false) @@ -297,11 +318,10 @@ describe('server-filesystem — real filesystem operations', () => { // Read back via MCP tool const readResult = await ctx.tools.execute({ - callId: nextCallId(), name: 'read_file', arguments: { path: filePath }, + callId: nextCallId(), name: 'mcp__filesystem__read_file', arguments: { path: filePath }, }) expect(readResult.isError).toBe(false) - const text = (readResult.content[0] as { type: string; text: string }).text - expect(text).toContain(content) + expect(textOf(readResult.content[0])).toContain(content) }) it('list_directory shows written file', async () => { @@ -309,10 +329,113 @@ describe('server-filesystem — real filesystem operations', () => { await writeFile(join(tempDir, 'listed.txt'), 'listed') const result = await ctx.tools.execute({ - callId: nextCallId(), name: 'list_directory', arguments: { path: tempDir }, + callId: nextCallId(), name: 'mcp__filesystem__list_directory', arguments: { path: tempDir }, }) expect(result.isError).toBe(false) - const text = (result.content[0] as { type: string; text: string }).text - expect(text).toContain('listed.txt') + expect(textOf(result.content[0])).toContain('listed.txt') + }) +}) + +// ---- Streamable HTTP transport ---- + +describe('streamable-http — in-process MCP server', () => { + let ctx: Context + let httpServer: Server + let baseUrl: string + /** Authorization header values observed by the HTTP server, in arrival order. */ + const seenAuth: Array = [] + + /** + * Stateless Streamable HTTP endpoint: a fresh McpServer + server transport + * per request (the SDK's documented stateless pattern — no session id, no + * SSE stream to keep). The tool set mirrors a minimal fixture server. + */ + async function handleMcpRequest(req: IncomingMessage, res: ServerResponse): Promise { + seenAuth.push(req.headers.authorization) + const server = new McpServer( + { name: 'http-fixture', version: '1.0.0' }, + { capabilities: { tools: {} } }, + ) + server.registerTool('ping', { + description: 'Replies pong.', + inputSchema: {}, + }, async () => ({ + content: [{ type: 'text', text: 'pong' }], + })) + server.registerTool('shout', { + description: 'Upper-cases a message.', + inputSchema: { message: z.string().describe('Message to upper-case') }, + }, async args => ({ + content: [{ type: 'text', text: args.message.toUpperCase() }], + })) + // Stateless mode: sessionIdGenerator ABSENT (the runtime treats absent and + // explicit-undefined identically; exactOptionalPropertyTypes forbids the + // SDK-documented explicit `sessionIdGenerator: undefined` spelling). + const transport = new StreamableHTTPServerTransport({}) + res.on('close', () => { void transport.close(); void server.close() }) + // Same exactOptionalPropertyTypes mismatch the client transport factory + // documents (src/transport.ts): the SDK types optional callbacks without + // `| undefined`. The SDK constructed the object; the cast is safe. + await server.connect(transport as Transport) + await transport.handleRequest(req, res) + } + + beforeAll(async () => { + httpServer = createServer((req, res) => { + handleMcpRequest(req, res).catch((error: unknown) => { + res.writeHead(500).end(String(error)) + }) + }) + const listening: PromiseWithResolvers = Promise.withResolvers() + httpServer.listen(0, '127.0.0.1', listening.resolve) + await listening.promise + const address = httpServer.address() + if (address === null || typeof address === 'string') throw new Error(`expected a TCP AddressInfo, got ${String(address)}`) + baseUrl = `http://127.0.0.1:${address.port}/mcp` + + ctx = await mountRegistry() + const config: Config = { + transport: 'streamable-http', + serverName: 'web', + url: baseUrl, + headers: { Authorization: 'Bearer e2e-test-token' }, + toolCallTimeoutMs: 15_000, + } + await applyAndWait(ctx, config) + }, 30_000) + + afterAll(async () => { + if (ctx) await ctx.fiber.dispose() + await sleep(200) + const closed: PromiseWithResolvers = Promise.withResolvers() + httpServer.close(() => { closed.resolve() }) + await closed.promise + }) + + it('discovers tools under the server namespace over HTTP', () => { + const names = ctx.tools.schemas().map(s => s.name) + expect(names).toContain('mcp__web__ping') + expect(names).toContain('mcp__web__shout') + }) + + it('executes ping() → "pong" over HTTP', async () => { + const result = await ctx.tools.execute({ + callId: nextCallId(), name: 'mcp__web__ping', arguments: {}, + }) + expect(result.isError).toBe(false) + expect(result.content[0]).toEqual({ type: 'text', text: 'pong' }) + }) + + it('executes shout({ message }) with args over HTTP', async () => { + const result = await ctx.tools.execute({ + callId: nextCallId(), name: 'mcp__web__shout', arguments: { message: 'quiet' }, + }) + expect(result.isError).toBe(false) + expect(result.content[0]).toEqual({ type: 'text', text: 'QUIET' }) + }) + + it('sends configured headers on every HTTP request', () => { + expect(seenAuth.length).toBeGreaterThan(0) + for (const auth of seenAuth) expect(auth).toBe('Bearer e2e-test-token') }) }) diff --git a/packages/mcp/mcp-client/tests/mcp-client.spec.ts b/packages/mcp/mcp-client/tests/mcp-client.spec.ts index e81369c0f0..8fff832434 100644 --- a/packages/mcp/mcp-client/tests/mcp-client.spec.ts +++ b/packages/mcp/mcp-client/tests/mcp-client.spec.ts @@ -3,7 +3,7 @@ import { Context } from 'cordis' import { CallId } from '@deepseek-ai/dsh-llm' import SystemPrompt from '@deepseek-ai/dsh-system-prompt' import ToolRegistry from '@deepseek-ai/dsh-tools' -import { syncTools, type ToolBridgeOptions } from '@deepseek-ai/dsh-mcp-client/src/tools.ts' +import { publicToolName, syncTools, type ToolBridgeOptions } from '@deepseek-ai/dsh-mcp-client/src/tools.ts' import { createTransport } from '@deepseek-ai/dsh-mcp-client/src/transport.ts' import type { Config } from '@deepseek-ai/dsh-mcp-client' @@ -40,12 +40,41 @@ async function mountRegistry(): Promise { } const defaultOpts: ToolBridgeOptions = { - toolPrefix: '', + serverName: 'srv', toolCallTimeoutMs: 60_000, } // ---- Tests ---- +describe('publicToolName', () => { + it('joins clean names verbatim', () => { + expect(publicToolName('github', 'create_issue')).toBe('mcp__github__create_issue') + expect(publicToolName('everything', 'get-sum')).toBe('mcp__everything__get-sum') + }) + + it('replaces invalid characters and appends an identity hash', () => { + const name = publicToolName('srv', 'admin.reset') + expect(name).toMatch(/^mcp__srv__admin_reset_[0-9a-f]{12}$/) + expect(name.length).toBeLessThanOrEqual(64) + }) + + it('truncates over-long names and appends an identity hash', () => { + const rawName = 'a'.repeat(80) + const name = publicToolName('srv', rawName) + expect(name).toHaveLength(64) + expect(name).toMatch(/_[0-9a-f]{12}$/) + expect(name.startsWith('mcp__srv__aaa')).toBe(true) + }) + + it('is deterministic and collision-free for distinct identities', () => { + // Two raw names that normalize to the same base must not collapse. + const a = publicToolName('srv', 'admin.reset') + const b = publicToolName('srv', 'admin_reset') + expect(a).toBe(publicToolName('srv', 'admin.reset')) + expect(a).not.toBe(b) + }) +}) + describe('syncTools', () => { let ctx: Context @@ -53,7 +82,7 @@ describe('syncTools', () => { ctx = await mountRegistry() }) - it('registers tools from listTools response', async () => { + it('registers tools under server-qualified public names', async () => { const client = createMockClient([ { name: 'greet', description: 'Say hello', inputSchema: { type: 'object', properties: { name: { type: 'string' } } } }, { name: 'add', description: 'Add numbers', inputSchema: { type: 'object', properties: {} } }, @@ -62,44 +91,85 @@ describe('syncTools', () => { const disposers = await syncTools(client as never, ctx, defaultOpts, new Map()) expect(disposers.size).toBe(2) - expect(ctx.tools.get('greet')).toBeDefined() - expect(ctx.tools.get('add')).toBeDefined() + expect(ctx.tools.get('mcp__srv__greet')).toBeDefined() + expect(ctx.tools.get('mcp__srv__add')).toBeDefined() + // Raw names are NOT registered. + expect(ctx.tools.get('greet')).toBeUndefined() + expect(ctx.tools.get('add')).toBeUndefined() }) - it('applies toolPrefix to registered names', async () => { - const client = createMockClient([ - { name: 'create_issue', description: 'Create an issue', inputSchema: { type: 'object' } }, - ]) + it('lets two servers publish the same raw name side by side', async () => { + const clientA = createMockClient([{ name: 'search', inputSchema: { type: 'object' } }]) + const clientB = createMockClient([{ name: 'search', inputSchema: { type: 'object' } }]) - const disposers = await syncTools(client as never, ctx, { ...defaultOpts, toolPrefix: 'gh_' }, new Map()) + await syncTools(clientA as never, ctx, { ...defaultOpts, serverName: 'github' }, new Map()) + await syncTools(clientB as never, ctx, { ...defaultOpts, serverName: 'web' }, new Map()) - expect(disposers.size).toBe(1) - expect(ctx.tools.get('gh_create_issue')).toBeDefined() - expect(ctx.tools.get('create_issue')).toBeUndefined() + expect(ctx.tools.get('mcp__github__search')).toBeDefined() + expect(ctx.tools.get('mcp__web__search')).toBeDefined() }) - it('skips tools with conflicting names and logs warning', async () => { - // Pre-register a tool with the same name. + it('coexists with a native tool of the same raw name', async () => { ctx.tools.register({ - name: 'existing', - description: 'Already here', + name: 'search', + description: 'Native search', parameters: { type: 'object' }, execute: async () => [{ type: 'text', text: 'native' }], }) + const client = createMockClient([{ name: 'search', inputSchema: { type: 'object' } }]) + await syncTools(client as never, ctx, defaultOpts, new Map()) + + expect(ctx.tools.get('search')).toBeDefined() + expect(ctx.tools.get('mcp__srv__search')).toBeDefined() + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'search', arguments: {} }) + expect(result.content[0]).toEqual({ type: 'text', text: 'native' }) + }) + + it('rejects a tool list where one raw name appears twice', async () => { const client = createMockClient([ - { name: 'existing', description: 'Conflicts', inputSchema: { type: 'object' } }, - { name: 'unique', description: 'No conflict', inputSchema: { type: 'object' } }, + { name: 'dup', inputSchema: { type: 'object' } }, + { name: 'dup', inputSchema: { type: 'object' } }, + ]) + + await expect(syncTools(client as never, ctx, defaultOpts, new Map())) + .rejects.toThrow(/listed tool "dup" more than once/) + // Nothing registered, previous generation untouched (it was empty). + expect(ctx.tools.get('mcp__srv__dup')).toBeUndefined() + }) + + it('keeps the previous generation when the fetch phase fails', async () => { + const client = createMockClient([{ name: 'stable', inputSchema: { type: 'object' } }]) + const first = await syncTools(client as never, ctx, defaultOpts, new Map()) + expect(ctx.tools.get('mcp__srv__stable')).toBeDefined() + + client.listTools.mockRejectedValue(new Error('network down')) + await expect(syncTools(client as never, ctx, defaultOpts, first)).rejects.toThrow('network down') + + // The previous generation is still live. + expect(ctx.tools.get('mcp__srv__stable')).toBeDefined() + }) + + it('rolls back the whole generation when a foreign tool squats on the namespace', async () => { + // A foreign registration occupies one of this server's public names. + ctx.tools.register({ + name: 'mcp__srv__taken', + description: 'Squatter', + parameters: { type: 'object' }, + execute: async () => [{ type: 'text', text: 'squatter' }], + }) + const client = createMockClient([ + { name: 'free', inputSchema: { type: 'object' } }, + { name: 'taken', inputSchema: { type: 'object' } }, ]) const disposers = await syncTools(client as never, ctx, defaultOpts, new Map()) - // Only the non-conflicting tool registers. - expect(disposers.size).toBe(1) - expect(ctx.tools.get('unique')).toBeDefined() - // Original tool unchanged. - const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'existing', arguments: {} }) - expect(result.content[0]).toEqual({ type: 'text', text: 'native' }) + // All-or-nothing: the non-conflicting tool is rolled back too. + expect(disposers.size).toBe(0) + expect(ctx.tools.get('mcp__srv__free')).toBeUndefined() + // The squatter is untouched. + expect(ctx.tools.get('mcp__srv__taken')).toBeDefined() }) it('unregisters previous tools before re-syncing', async () => { @@ -108,14 +178,14 @@ describe('syncTools', () => { ]) const firstDisposers = await syncTools(client as never, ctx, defaultOpts, new Map()) - expect(ctx.tools.get('old_tool')).toBeDefined() + expect(ctx.tools.get('mcp__srv__old_tool')).toBeDefined() // Second sync with different tools should remove old_tool. client.listTools.mockResolvedValue({ tools: [{ name: 'new_tool', inputSchema: { type: 'object' } }], nextCursor: undefined }) const secondDisposers = await syncTools(client as never, ctx, defaultOpts, firstDisposers) - expect(ctx.tools.get('old_tool')).toBeUndefined() - expect(ctx.tools.get('new_tool')).toBeDefined() + expect(ctx.tools.get('mcp__srv__old_tool')).toBeUndefined() + expect(ctx.tools.get('mcp__srv__new_tool')).toBeDefined() expect(secondDisposers.size).toBe(1) }) @@ -128,8 +198,8 @@ describe('syncTools', () => { const disposers = await syncTools(client as never, ctx, defaultOpts, new Map()) expect(disposers.size).toBe(2) - expect(ctx.tools.get('page1')).toBeDefined() - expect(ctx.tools.get('page2')).toBeDefined() + expect(ctx.tools.get('mcp__srv__page1')).toBeDefined() + expect(ctx.tools.get('mcp__srv__page2')).toBeDefined() }) }) @@ -140,17 +210,18 @@ describe('tool execution', () => { ctx = await mountRegistry() }) - it('calls MCP callTool and returns text content', async () => { + it('calls MCP callTool with the RAW name and returns text content', async () => { const client = createMockClient( [{ name: 'echo', inputSchema: { type: 'object' } }], { content: [{ type: 'text', text: 'hello world' }] }, ) await syncTools(client as never, ctx, defaultOpts, new Map()) - const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'echo', arguments: { msg: 'hi' } }) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'mcp__srv__echo', arguments: { msg: 'hi' } }) expect(result.isError).toBe(false) expect(result.content).toEqual([{ type: 'text', text: 'hello world' }]) + // The wire sees the raw MCP name, never the public name. expect(client.callTool).toHaveBeenCalledWith( { name: 'echo', arguments: { msg: 'hi' } }, undefined, @@ -158,6 +229,24 @@ describe('tool execution', () => { ) }) + it('sends the raw name for normalized public names', async () => { + const client = createMockClient( + [{ name: 'admin.reset', inputSchema: { type: 'object' } }], + { content: [{ type: 'text', text: 'reset done' }] }, + ) + + await syncTools(client as never, ctx, defaultOpts, new Map()) + const publicName = publicToolName('srv', 'admin.reset') + const result = await ctx.tools.execute({ callId: CallId('c1'), name: publicName, arguments: {} }) + + expect(result.isError).toBe(false) + expect(client.callTool).toHaveBeenCalledWith( + { name: 'admin.reset', arguments: {} }, + undefined, + expect.anything(), + ) + }) + it('joins multiple text blocks with newline', async () => { const client = createMockClient( [{ name: 'multi', inputSchema: { type: 'object' } }], @@ -165,7 +254,7 @@ describe('tool execution', () => { ) await syncTools(client as never, ctx, defaultOpts, new Map()) - const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'multi', arguments: {} }) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'mcp__srv__multi', arguments: {} }) expect(result.content).toEqual([{ type: 'text', text: 'line1\nline2' }]) }) @@ -177,7 +266,7 @@ describe('tool execution', () => { ) await syncTools(client as never, ctx, defaultOpts, new Map()) - const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'img', arguments: {} }) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'mcp__srv__img', arguments: {} }) expect(result.content[0]).toEqual({ type: 'text', text: 'before\n[image: image/png, content discarded]' }) }) @@ -189,7 +278,7 @@ describe('tool execution', () => { ) await syncTools(client as never, ctx, defaultOpts, new Map()) - const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'fail', arguments: {} }) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'mcp__srv__fail', arguments: {} }) expect(result.isError).toBe(true) expect(result.content[0]).toEqual({ type: 'text', text: 'Error: something went wrong' }) @@ -203,7 +292,7 @@ describe('tool execution', () => { ) await syncTools(client as never, ctx, defaultOpts, new Map()) - await ctx.tools.execute({ callId: CallId('c1'), name: 'slow', arguments: {}, signal: controller.signal }) + await ctx.tools.execute({ callId: CallId('c1'), name: 'mcp__srv__slow', arguments: {}, signal: controller.signal }) expect(client.callTool).toHaveBeenCalledWith( expect.anything(), @@ -219,7 +308,7 @@ describe('tool execution', () => { client.callTool.mockResolvedValue({ toolResult: { key: 'value' } }) await syncTools(client as never, ctx, defaultOpts, new Map()) - const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'legacy', arguments: {} }) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'mcp__srv__legacy', arguments: {} }) expect(result.isError).toBe(false) expect(result.content[0]).toEqual({ type: 'text', text: '{"key":"value"}' }) @@ -240,7 +329,7 @@ describe('tool execution edge cases', () => { ) await syncTools(client as never, ctx, defaultOpts, new Map()) - const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'audio_tool', arguments: {} }) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'mcp__srv__audio_tool', arguments: {} }) expect(result.content[0]).toEqual({ type: 'text', text: '[audio: audio/mp3, content discarded]' }) }) @@ -252,7 +341,7 @@ describe('tool execution edge cases', () => { ) await syncTools(client as never, ctx, defaultOpts, new Map()) - const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'res_tool', arguments: {} }) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'mcp__srv__res_tool', arguments: {} }) expect(result.content[0]).toEqual({ type: 'text', text: '[resource: content discarded]' }) }) @@ -264,7 +353,7 @@ describe('tool execution edge cases', () => { ) await syncTools(client as never, ctx, defaultOpts, new Map()) - const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'link_tool', arguments: {} }) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'mcp__srv__link_tool', arguments: {} }) expect(result.content[0]).toEqual({ type: 'text', text: '[resource: content discarded]' }) }) @@ -276,7 +365,7 @@ describe('tool execution edge cases', () => { ) await syncTools(client as never, ctx, defaultOpts, new Map()) - const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'unknown_tool', arguments: {} }) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'mcp__srv__unknown_tool', arguments: {} }) expect(result.content[0]).toEqual({ type: 'text', text: '[unsupported content type: video]' }) }) @@ -288,7 +377,7 @@ describe('tool execution edge cases', () => { ) await syncTools(client as never, ctx, defaultOpts, new Map()) - const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'img2', arguments: {} }) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'mcp__srv__img2', arguments: {} }) expect(result.content[0]).toEqual({ type: 'text', text: '[image: unknown, content discarded]' }) }) @@ -300,7 +389,7 @@ describe('tool execution edge cases', () => { ) await syncTools(client as never, ctx, defaultOpts, new Map()) - const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'audio_no_mime', arguments: {} }) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'mcp__srv__audio_no_mime', arguments: {} }) expect(result.content[0]).toEqual({ type: 'text', text: '[audio: unknown, content discarded]' }) }) @@ -312,7 +401,7 @@ describe('tool execution edge cases', () => { ) await syncTools(client as never, ctx, defaultOpts, new Map()) - const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'notext', arguments: {} }) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'mcp__srv__notext', arguments: {} }) expect(result.content[0]).toEqual({ type: 'text', text: '(notext returned no text content)' }) }) @@ -324,7 +413,7 @@ describe('tool execution edge cases', () => { ) await syncTools(client as never, ctx, defaultOpts, new Map()) - const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'empty_tool', arguments: {} }) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'mcp__srv__empty_tool', arguments: {} }) expect(result.content[0]).toEqual({ type: 'text', text: '(empty_tool returned no text content)' }) }) @@ -337,7 +426,7 @@ describe('tool execution edge cases', () => { client.callTool.mockResolvedValue({}) await syncTools(client as never, ctx, defaultOpts, new Map()) - const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'legacy2', arguments: {} }) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'mcp__srv__legacy2', arguments: {} }) expect(result.content[0]).toEqual({ type: 'text', text: '(no output)' }) }) @@ -349,13 +438,9 @@ describe('tool execution edge cases', () => { ) await syncTools(client as never, ctx, defaultOpts, new Map()) - const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'err_notext', arguments: {} }) + const result = await ctx.tools.execute({ callId: CallId('c1'), name: 'mcp__srv__err_notext', arguments: {} }) expect(result.isError).toBe(true) - // The error message falls back to 'MCP tool error' when content[0] is not text. - // But mapContent converts image to text placeholder, so it should use that. - // Actually mapContent ALWAYS returns text, so the ternary always takes the truthy branch. - // Let me check: mapContent returns [{type:'text', text:'[image: ...]'}], so content[0].type IS 'text'. expect(result.content[0]).toEqual({ type: 'text', text: 'Error: [image: image/png, content discarded]' }) }) @@ -366,7 +451,7 @@ describe('tool execution edge cases', () => { ]) await syncTools(client as never, ctx, defaultOpts, new Map()) - const tool = ctx.tools.get('described') + const tool = ctx.tools.get('mcp__srv__described') expect(tool?.description).toBe('A described tool') }) @@ -376,7 +461,7 @@ describe('tool execution edge cases', () => { ]) await syncTools(client as never, ctx, defaultOpts, new Map()) - const tool = ctx.tools.get('nodesc') + const tool = ctx.tools.get('mcp__srv__nodesc') expect(tool?.description).toBe('') }) }) @@ -385,11 +470,11 @@ describe('createTransport', () => { it('creates StdioClientTransport for stdio config', () => { const config: Config = { transport: 'stdio', + serverName: 'srv', command: 'node', args: ['server.js'], env: {}, cwd: '/tmp', - toolPrefix: '', toolCallTimeoutMs: 60_000, } const transport = createTransport(config) @@ -401,9 +486,9 @@ describe('createTransport', () => { it('creates StreamableHTTPClientTransport for http config without headers', () => { const config: Config = { transport: 'streamable-http', + serverName: 'srv', url: 'http://localhost:3000/mcp', headers: {}, - toolPrefix: '', toolCallTimeoutMs: 60_000, } const transport = createTransport(config) @@ -415,9 +500,9 @@ describe('createTransport', () => { it('creates StreamableHTTPClientTransport for http config with headers', () => { const config: Config = { transport: 'streamable-http', + serverName: 'srv', url: 'http://localhost:3000/mcp', headers: { Authorization: 'Bearer token' }, - toolPrefix: '', toolCallTimeoutMs: 60_000, } const transport = createTransport(config) @@ -436,11 +521,11 @@ describe('createTransport', () => { const config: Config = { transport: 'stdio', + serverName: 'srv', command: 'echo', args: [], env: { EXTRA: 'injected' }, cwd: '', - toolPrefix: '', toolCallTimeoutMs: 60_000, } // createTransport internally calls buildChildEnv; we verify by inspecting @@ -463,11 +548,11 @@ describe('createTransport', () => { it('merges explicit env on top of scrubbed ambient env', () => { const config: Config = { transport: 'stdio', + serverName: 'srv', command: 'echo', args: [], env: { CUSTOM: 'value' }, cwd: '', - toolPrefix: '', toolCallTimeoutMs: 60_000, } const transport = createTransport(config) @@ -490,7 +575,7 @@ describe('tool execution — non-object args fallback', () => { await syncTools(client as never, ctx, defaultOpts, new Map()) // Simulate model emitting `null` as tool arguments (malformed). - await ctx.tools.execute({ callId: CallId('c1'), name: 'coerce', arguments: null }) + await ctx.tools.execute({ callId: CallId('c1'), name: 'mcp__srv__coerce', arguments: null }) expect(client.callTool).toHaveBeenCalledWith( { name: 'coerce', arguments: {} }, @@ -506,7 +591,7 @@ describe('tool execution — non-object args fallback', () => { ) await syncTools(client as never, ctx, defaultOpts, new Map()) - await ctx.tools.execute({ callId: CallId('c1'), name: 'coerce2', arguments: 'bad' }) + await ctx.tools.execute({ callId: CallId('c1'), name: 'mcp__srv__coerce2', arguments: 'bad' }) expect(client.callTool).toHaveBeenCalledWith( { name: 'coerce2', arguments: {} }, @@ -515,4 +600,3 @@ describe('tool execution — non-object args fallback', () => { ) }) }) - From 01da49a3aba3ee285784262af847988f70935ad9 Mon Sep 17 00:00:00 2001 From: Tianyi Cui <53024+tianyicui@users.noreply.github.com> Date: Tue, 14 Jul 2026 03:07:41 +0800 Subject: [PATCH 21/26] refactor: prune code runtime surface --- docs/cordis-catalog/services.md | 2 +- docs/core-data-structures/code-runtime.md | 17 ++----- .../feature/2026-06-15-code-mode.md | 5 +- .../code-runtime-worker/README.md | 4 +- .../code-runtime-worker/package.json | 1 - .../code-runtime-worker/src/bootstrap.ts | 40 +++++++-------- .../code-runtime-worker/src/index.ts | 44 ++++++---------- .../code-runtime-worker/src/protocol.ts | 8 ++- .../tests/bootstrap.spec.ts | 50 +++++++++---------- .../tests/built-lib.e2e.ts | 4 +- .../code-runtime-worker/tests/runtime.spec.ts | 36 ++++++------- packages/code-runtime/code-runtime/README.md | 2 +- .../code-runtime/code-runtime/src/index.ts | 1 - .../code-runtime/code-runtime/src/types.ts | 18 +------ .../code-runtime/tests/service.spec.ts | 2 +- .../cordis/tool-cordis/src/api-catalog.ts | 6 +-- packages/core/tools/src/code-mode.ts | 11 ++-- packages/core/tools/tests/code-mode.spec.ts | 11 ++-- scripts/type-equiv.manifest.json | 1 - 19 files changed, 100 insertions(+), 163 deletions(-) diff --git a/docs/cordis-catalog/services.md b/docs/cordis-catalog/services.md index 28d1904ecf..e95ad6af70 100644 --- a/docs/cordis-catalog/services.md +++ b/docs/cordis-catalog/services.md @@ -98,7 +98,7 @@ abstract run(request: CodeRunRequest): Promise Types: [CodeRunRequest](../core-data-structures/code-runtime.md) · [CodeRunResult](../core-data-structures/code-runtime.md) -Source: [`packages/code-runtime/code-runtime/src/index.ts:59`](../../packages/code-runtime/code-runtime/src/index.ts) +Source: [`packages/code-runtime/code-runtime/src/index.ts:58`](../../packages/code-runtime/code-runtime/src/index.ts) ## `ctx.compact` — `CompactService` (abstract seam) diff --git a/docs/core-data-structures/code-runtime.md b/docs/core-data-structures/code-runtime.md index cb1661e02f..9237a3cce9 100644 --- a/docs/core-data-structures/code-runtime.md +++ b/docs/core-data-structures/code-runtime.md @@ -39,8 +39,8 @@ interface CodeRunResult { * or value-less run leaves this absent. */ value?: unknown - /** Everything the program emitted, in order (capped by the implementation). */ - logs: CodeLogEntry[] + /** Text the program emitted, in order (capped by the implementation). */ + logs: string[] /** Present iff the run failed; see {@link CodeRunFailure} for the taxonomy. */ error?: CodeRunFailure } @@ -65,18 +65,7 @@ type CodeBindingFunction = (args: unknown) => Promise ## Captured output and the failure taxonomy -Logs arrive in emission order, attributed to their channel (the runtime's `console` shim, or stray writes to the underlying streams): - -```ts type-equiv -interface CodeLogEntry { - /** Which channel produced the text. */ - source: 'console' | 'stdout' | 'stderr' - /** The console method used; present only when `source` is `'console'`. */ - level?: 'log' | 'info' | 'warn' | 'error' | 'debug' - /** The captured text (possibly truncated by the implementation's caps, marked in-band). */ - text: string -} -``` +Logs are plain strings in emission order. The runtime captures the program's console and stream output, but channel and console-method metadata are not part of the seam because consumers render only the text. Implementations cap the aggregate output and mark truncation in-band. Failure kinds are **orthogonal outcomes reported independently** (per [defensive-patterns](../defensive-patterns.md)): a budget expiry is not an exception, an abort is not a timeout, and a substrate death (e.g. OOM) is neither: diff --git a/docs/rfc/implemented/feature/2026-06-15-code-mode.md b/docs/rfc/implemented/feature/2026-06-15-code-mode.md index 60acfa7c88..32d60aafcf 100644 --- a/docs/rfc/implemented/feature/2026-06-15-code-mode.md +++ b/docs/rfc/implemented/feature/2026-06-15-code-mode.md @@ -40,7 +40,7 @@ Under `'code'` and `'both'` the registry owns `run_code` as a reserved presentat 1. **Builds the bindings**: the bridge owns a **run-scoped `AbortController`** whose signal follows `exec.signal` (an outer cancel propagates in) and which the bridge itself fires the moment the run settles for any reason — completion, program exception, `computeMs`/`maxWallMs` expiry, worker exit. For every visible capability tool, the binding is an async function that (a) checks the run signal before and after, (b) **JSON-normalizes the argument** — a `JSON.parse(JSON.stringify(args))` round-trip, rejecting that one call with a descriptive `Error` when the value does not survive (`BigInt`, circular structures) — because the seam's structured-clone boundary is wider than JSON while the session log accepts only JSON, (c) awaits its turn on the **per-run serialization queue** (below), (d) calls `this.execute({ callId, name, arguments, agent: exec.agent, parent: exec.token, signal: runSignal })` with a deterministic sub-id `` CallId(`${exec.callId}:code:${n}`) ``, (e) appends a `tool/code-dispatch` session event, and (f) maps the result: success → the text-block contents joined as a `string` (non-text blocks become placeholders), `isError` → **the binding rejects** with an `Error` carrying the result text. The child's readonly `parent` is only the outer execution's frozen, property-free token, so commit-style observers can correlate outcomes without receiving a mutation path into the live `run_code` wrapper. Every sub-call still traverses the full pipeline under its own immutable identity and registry-assigned token. The run signal, rather than the bare outer one, lets budget expiry abort an in-flight sub-tool instead of orphaning it. Rejection gives programs ordinary `try/catch` and `Promise.all` failure semantics rather than a bespoke result envelope. 2. **Runs the program**: `ctx.codeRuntime.run({ program: args.code, bindings: [{ global: 'tools', functions }], signal: runController.signal })`. The runtime receives the run-scoped signal, not only the caller's outer signal, so any way the outer run settles also aborts work inside the runtime. -3. **Surfaces the outcome — after reaching quiescence.** When `ctx.codeRuntime.run()` settles, whether by fulfillment or rejection, the bridge fires the run-scoped abort (cancelling any in-flight sub-dispatch and abandoning queued-unstarted ones), then **awaits the dispatch queue's drain before returning or propagating**, per the dispose-to-quiescence rule in [defensive patterns](../../../defensive-patterns.md): an aborted in-flight sub-call still settles and logs its `isError` `tool/code-dispatch` event *inside* the open turn, and nothing can append after `run_code` settles. A successful result then returns one text block — the captured console/stdout output followed by the rendered return value (if any) — plus a `meta` payload (capped logs, dispatch count) for presentation. A fulfilled run with `result.error` throws a `CodeRunFailedError extends HarnessError` (`code: 'CODE_RUN_FAILED'`, message = the error kind and text plus captured logs so the model can self-correct); a backend rejection propagates through the same registry error boundary. Both become structured `isError` tool results. +3. **Surfaces the outcome — after reaching quiescence.** When `ctx.codeRuntime.run()` settles, whether by fulfillment or rejection, the bridge fires the run-scoped abort (cancelling any in-flight sub-dispatch and abandoning queued-unstarted ones), then **awaits the dispatch queue's drain before returning or propagating**, per the dispose-to-quiescence rule in [defensive patterns](../../../defensive-patterns.md): an aborted in-flight sub-call still settles and logs its `isError` `tool/code-dispatch` event *inside* the open turn, and nothing can append after `run_code` settles. A successful result then returns one text block — the captured console/stdout output followed by the rendered return value (if any) — plus the capped log strings as presentation metadata. A fulfilled run with `result.error` throws a `CodeRunFailedError extends HarnessError` (`code: 'CODE_RUN_FAILED'`, message = the error kind and text plus captured logs so the model can self-correct); a backend rejection propagates through the same registry error boundary. Both become structured `isError` tool results. **Sub-call `additionalContext` is suppressed, deliberately.** A `tools/post-execute` hook may attach `additionalContext` to a call; for loop-dispatched calls the loop buffers those and appends each as a `context/message` only after the step's `tool/result`s, preserving call/result adjacency. A sub-dispatch result's `additionalContext` has no such safe outlet from inside a running `run_code`: injecting immediately would land a `context/message` between the parent's `tool/call` and its `tool/result` (breaking the adjacency the buffering exists to protect), and `PostToolDecision.additionalContext` is singular where a program may produce many. The MVP therefore drops sub-call `additionalContext`, pinned by a test and stated in the hooks bridge's docs; the follow-up (a plural context channel or loop-level sub-dispatch buffering) is deferred until a real hook needs it through Code Mode. @@ -58,8 +58,7 @@ Each sub-dispatch appends one session event, declared by `dsh-tools` via `Sessio - `CodeRunRequest = { program: string; bindings: CodeBindingNamespace[]; signal?: AbortSignal }` - `CodeBindingNamespace = { global: string; functions: Record Promise> }` — the runtime exposes each namespace as a global object of async functions inside the program; binding arguments and resolutions must be structured-cloneable (a runtime may cross a serialization boundary; ours does). -- `CodeRunResult = { value?: unknown; logs: CodeLogEntry[]; error?: CodeRunFailure }` — program execution outcomes, including exception, timeout, abort, and worker exit, resolve as the `error` field. `run()` may reject only for caller/seam misuse (for example a duplicate binding namespace); consumers still contain a non-conforming backend rejection at their own error boundary. -- `CodeLogEntry = { source: 'console' | 'stdout' | 'stderr'; level?: 'log' | 'info' | 'warn' | 'error' | 'debug'; text: string }` +- `CodeRunResult = { value?: unknown; logs: string[]; error?: CodeRunFailure }` — program execution outcomes, including exception, timeout, abort, and worker exit, resolve as the `error` field. `run()` may reject only for caller/seam misuse (for example a duplicate binding namespace); consumers still contain a non-conforming backend rejection at their own error boundary. - `CodeRunFailure = { kind: 'exception' | 'timeout' | 'abort' | 'worker-exit'; message: string }` — orthogonal outcomes reported independently per [defensive patterns](../../../defensive-patterns.md); a timed-out run is not an exception, an abort is not a timeout. - Two readonly backend descriptors, informational not gating: `language` (what the program must be written in — `'typescript'` for the shipped backend; a Python backend would say so, and pair with its own SDK generator on the presentation side) and `isolation` (`'worker-thread'` for the shipped backend; `'process'`, `'container'`, … for future ones). `dsh-tools` requires `language === 'typescript'` in the MVP — its codegen emits TS — and fails the assembly loudly otherwise, the same misconfiguration idiom as `toolOrder` violations (as when `mode` is non-native with no `ctx.codeRuntime` loaded at all). diff --git a/packages/code-runtime/code-runtime-worker/README.md b/packages/code-runtime/code-runtime-worker/README.md index 8138f01539..705174d0e5 100644 --- a/packages/code-runtime/code-runtime-worker/README.md +++ b/packages/code-runtime/code-runtime-worker/README.md @@ -23,10 +23,12 @@ Every field is validated (positive numbers) and defaulted; there are no other tu - **Type-strip host-side, in execution context** — the program is wrapped in an async-function shell, stripped with `node:module`'s `stripTypeScriptTypes` (erasable syntax only — `enum`/namespaces are rejected as a program `exception` and no worker spawns), and sliced back out byte-positioned; it then executes as the body of an `AsyncFunction`, so top-level `await`/`return` work. - **The port assumes a hostile peer** — model code can reach `parentPort` and forge traffic, so every inbound message is shape-validated and REBUILT before anything reads it (`null`, primitives, junk types, and malformed payloads drop without a throw; forged extra fields never ride along), the host answers each call id at most once, resolves binding names as OWN properties only (a forged `constructor` cannot walk a prototype chain), drops post-settlement replies, and converts a non-cloneable binding resolution into an error reply. Forged `log`/`done` messages cannot bypass the caps: one host-side ledger bounds everything that lands in `logs`, and the completion value is re-capped host-side. Worker-side namespaces are null-prototype with `defineProperty`, so `__proto__`-shaped binding names are ordinary keys. - **Two independent budgets, because the peer is hostile** — `computeMs` meters the worker's MEASURED busy time (`worker.performance.eventLoopUtilization()` polling): a hot loop cannot hide behind a pending decoy dispatch, and a program awaiting a slow tool accrues nothing. `maxWallMs` backstops what busy time cannot see (awaiting a promise nobody resolves). Both funnel into `worker.terminate()`, which ends hot synchronous loops too; heap overflow surfaces as the worker's OOM exit (`kind: 'worker-exit'`). -- **Logs stream eagerly** — console/stdout/stderr entries cross the port as they happen, so a timed-out or killed program still shows what it printed. ONE shared `maxLogBytes` ledger bounds everything: streamed entries, forged port traffic, and pipe bytes that bypass the patched streams (appended after), with the overflow marked in-band once. +- **Logs stream eagerly** — console/stdout/stderr text crosses the port in emission order, so a timed-out or killed program still shows what it printed. ONE shared `maxLogBytes` ledger bounds everything: streamed text, forged port traffic, and pipe bytes that bypass the patched streams (appended after), with the overflow marked in-band once. - **Empty environment** — the worker gets `env: {}` and `execArgv: []`: no ambient credentials (stronger than the scrubbed-env rule for spawned commands) and no inherited loader flags. - **Dispose to quiescence** — teardown fails in-flight runs as `abort` and AWAITS each worker's exit before resolving. ## The worker entry, unbuilt and built `worker.ts` is deliberately erasable-only TypeScript with type-only cross-package imports: unbuilt (vitest/tsx), the host spawns `src/worker.ts` directly and Node's native type stripping loads it; built, the entry ships as the sibling CommonJS bundle `lib/worker.cjs` (its own tsdown entry). The CommonJS format is required because pkg's VFS Worker hook compiles filesystem-string entries as CommonJS. The host converts either entry URL to a filesystem string before constructing `Worker`, which works through both ordinary Node resolution and that pkg hook. The built path is pinned by `tests/built-lib.e2e.ts`, the real-load-path guard from [docs/testing.md](../../../docs/testing.md). + +The SDK surface is the default/named `WorkerCodeRuntime` class plus `Config`. The operational `./worker` subpath exists only as the packaged spawn entry; the wire protocol and bootstrap helpers are source-private implementation details. diff --git a/packages/code-runtime/code-runtime-worker/package.json b/packages/code-runtime/code-runtime-worker/package.json index 91075243d2..3c4ee1d931 100644 --- a/packages/code-runtime/code-runtime-worker/package.json +++ b/packages/code-runtime/code-runtime-worker/package.json @@ -15,7 +15,6 @@ "types": "./lib/types/worker.d.ts", "default": "./lib/worker.cjs" }, - "./src/*": "./src/*", "./package.json": "./package.json" }, "files": [ diff --git a/packages/code-runtime/code-runtime-worker/src/bootstrap.ts b/packages/code-runtime/code-runtime-worker/src/bootstrap.ts index f2e0d343f3..a152b6b278 100644 --- a/packages/code-runtime/code-runtime-worker/src/bootstrap.ts +++ b/packages/code-runtime/code-runtime-worker/src/bootstrap.ts @@ -12,7 +12,6 @@ import { inspect } from 'node:util' import { serialize } from 'node:v8' -import type { CodeLogEntry } from '@deepseek-ai/dsh-code-runtime' import { logTruncationMarker } from './protocol.ts' import type { DoneMessage, ReplyMessage, WorkerBootData, WorkerToHost } from './protocol.ts' @@ -33,12 +32,12 @@ export interface PatchableStream { } /** - * Ordered log capture under one shared byte budget, delivered to a sink as - * each entry lands (the real sink streams entries over the port eagerly, so + * Ordered text capture under one shared byte budget, delivered to a sink as + * each item lands (the real sink streams text over the port eagerly, so * captured output survives a mid-run termination). Once the budget is - * exhausted it emits exactly one in-band marker entry (on the `stderr` - * diagnostics channel) and silently drops everything after — the cap is a - * blast-radius bound, so "how much was lost" intentionally stays unmeasured. + * exhausted it emits exactly one in-band marker and silently drops everything + * after. The cap is a blast-radius bound, so "how much was lost" intentionally + * stays unmeasured. */ export class LogBuffer { private remaining: number @@ -47,28 +46,28 @@ export class LogBuffer { // under Node's native strip-only mode, which rejects non-erasable syntax — // and parameter properties are non-erasable. private readonly maxBytes: number - private readonly sink: (entry: CodeLogEntry) => void + private readonly sink: (text: string) => void - constructor(maxBytes: number, sink: (entry: CodeLogEntry) => void) { + constructor(maxBytes: number, sink: (text: string) => void) { this.maxBytes = maxBytes this.sink = sink this.remaining = maxBytes } /** - * Emit one entry to the sink, charging its text against the budget (drops + marks once exhausted). - * @param entry - the log entry to deliver. + * Emit text to the sink, charging it against the budget (drops + marks once exhausted). + * @param text - the captured text to deliver. */ - push(entry: CodeLogEntry): void { + push(text: string): void { if (this.truncated) return - const cost = Buffer.byteLength(entry.text, 'utf8') + const cost = Buffer.byteLength(text, 'utf8') if (cost > this.remaining) { this.truncated = true - this.sink({ source: 'stderr', text: logTruncationMarker(this.maxBytes) }) + this.sink(logTruncationMarker(this.maxBytes)) return } this.remaining -= cost - this.sink(entry) + this.sink(text) } } @@ -89,7 +88,7 @@ export function makeConsoleShim(logs: LogBuffer): Record<(typeof CONSOLE_LEVELS) args.map(arg => typeof arg === 'string' ? arg : inspect(arg, INSPECT_OPTIONS)).join(' ') const shim = Object.create(null) as Record<(typeof CONSOLE_LEVELS)[number], (...args: unknown[]) => void> for (const level of CONSOLE_LEVELS) { - shim[level] = (...args: unknown[]) => { logs.push({ source: 'console', level, text: render(args) }) } + shim[level] = (...args: unknown[]) => { logs.push(render(args)) } } return shim } @@ -104,17 +103,16 @@ export function makeConsoleShim(logs: LogBuffer): Record<(typeof CONSOLE_LEVELS) * even for writes the exhausted budget drops. * @param logs - the buffer captured writes are pushed into. * @param stream - the stream whose `write` slot is patched. - * @param source - the log source the captured writes are attributed to. * @returns the restore function (the in-process tests un-patch; the real * worker never needs to). */ -export function captureStreamWrites(logs: LogBuffer, stream: PatchableStream, source: 'stdout' | 'stderr'): () => void { +export function captureStreamWrites(logs: LogBuffer, stream: PatchableStream): () => void { // The slot's VALUE is stored for restore and reassigned — never invoked // detached, so the unbound-method concern does not apply. // eslint-disable-next-line @typescript-eslint/unbound-method const original = stream.write stream.write = (chunk: unknown, ...rest: unknown[]): boolean => { - logs.push({ source, text: typeof chunk === 'string' ? chunk : String(chunk) }) + logs.push(typeof chunk === 'string' ? chunk : String(chunk)) // Node's optional-encoding shape: the callback is whichever of the next // two positions holds a function (a non-function there is the encoding). const callback = [rest[0], rest[1]].find( @@ -273,9 +271,9 @@ export async function runWorkerMain( data: WorkerBootData, streams: { stdout: PatchableStream; stderr: PatchableStream }, ): Promise { - const logs = new LogBuffer(data.maxLogBytes, (entry) => { port.postMessage({ type: 'log', entry }) }) - captureStreamWrites(logs, streams.stdout, 'stdout') - captureStreamWrites(logs, streams.stderr, 'stderr') + const logs = new LogBuffer(data.maxLogBytes, (text) => { port.postMessage({ type: 'log', text }) }) + captureStreamWrites(logs, streams.stdout) + captureStreamWrites(logs, streams.stderr) const pending = new Map() wireReplies(port, pending) diff --git a/packages/code-runtime/code-runtime-worker/src/index.ts b/packages/code-runtime/code-runtime-worker/src/index.ts index 7a8024cb3d..1f659ad769 100644 --- a/packages/code-runtime/code-runtime-worker/src/index.ts +++ b/packages/code-runtime/code-runtime-worker/src/index.ts @@ -18,7 +18,7 @@ import { fileURLToPath } from 'node:url' import { Context } from 'cordis' import z from 'schemastery' import { CodeRuntime } from '@deepseek-ai/dsh-code-runtime' -import type { CodeBindingFunction, CodeLogEntry, CodeRunFailure, CodeRunRequest, CodeRunResult } from '@deepseek-ai/dsh-code-runtime' +import type { CodeBindingFunction, CodeRunFailure, CodeRunRequest, CodeRunResult } from '@deepseek-ai/dsh-code-runtime' import { prepareValue, truncateUtf8Bytes } from './bootstrap.ts' import { logTruncationMarker } from './protocol.ts' import type { ReplyMessage, WorkerBootData, WorkerToHost } from './protocol.ts' @@ -118,10 +118,6 @@ function messageOf(error: unknown): string { return error instanceof Error ? error.message : String(error) } -/** The log sources / console levels the seam vocabulary admits, as runtime sets for inbound-message validation. */ -const LOG_SOURCES = new Set(['console', 'stdout', 'stderr']) -const LOG_LEVELS = new Set(['log', 'info', 'warn', 'error', 'debug']) - /** * Runtime shape gate for inbound port traffic. The peer runs MODEL CODE and * can post anything — `null`, primitives, objects with poisoned fields — so @@ -140,20 +136,8 @@ function parseWorkerMessage(raw: unknown): WorkerToHost | undefined { return { type: 'call', id: m.id, global: m.global, name: m.name, args: m.args } } case 'log': { - const entry = m.entry - if (typeof entry !== 'object' || entry === null) return undefined - const e = entry as Record - if (typeof e.text !== 'string') return undefined - if (typeof e.source !== 'string' || !LOG_SOURCES.has(e.source)) return undefined - if (e.level !== undefined && (typeof e.level !== 'string' || !LOG_LEVELS.has(e.level))) return undefined - return { - type: 'log', - entry: { - source: e.source as CodeLogEntry['source'], - ...e.level !== undefined ? { level: e.level as Exclude } : {}, - text: e.text, - }, - } + if (typeof m.text !== 'string') return undefined + return { type: 'log', text: m.text } } case 'done': { if (m.error === undefined) return { type: 'done', ...m.value !== undefined ? { value: m.value } : {} } @@ -299,8 +283,8 @@ export class WorkerCodeRuntime extends CodeRuntime { return new Promise((resolve) => { let settled = false const answered = new Set() - const logs: CodeLogEntry[] = [] - const strayLogs: CodeLogEntry[] = [] + const logs: string[] = [] + const strayLogs: string[] = [] // ONE host-side ledger for everything that lands in `logs`/`strayLogs`, // whatever the path: honest port entries, FORGED port entries (model @@ -310,26 +294,26 @@ export class WorkerCodeRuntime extends CodeRuntime { // so the documented cap is one shared `maxLogBytes` however it is hit. let logBudget = this.config.maxLogBytes let logsTruncated = false - const admit = (entry: CodeLogEntry, sink: CodeLogEntry[]): void => { + const admit = (text: string, sink: string[]): void => { if (logsTruncated) return - const cost = Buffer.byteLength(entry.text, 'utf8') + const cost = Buffer.byteLength(text, 'utf8') if (cost > logBudget) { logsTruncated = true - sink.push({ source: 'stderr', text: logTruncationMarker(this.config.maxLogBytes) }) + sink.push(logTruncationMarker(this.config.maxLogBytes)) return } logBudget -= cost - sink.push(entry) + sink.push(text) } // No settled guard: `finish` snapshots the arrays when it resolves, so // a chunk flushing after settlement mutates only the discarded buffers, // and the ledger bounds that growth until the pipes close. - const captureStray = (source: 'stdout' | 'stderr') => (chunk: Buffer) => { - admit({ source, text: chunk.toString('utf8') }, strayLogs) + const captureStray = (chunk: Buffer): void => { + admit(chunk.toString('utf8'), strayLogs) } - worker.stdout.on('data', captureStray('stdout')) - worker.stderr.on('data', captureStray('stderr')) + worker.stdout.on('data', captureStray) + worker.stderr.on('data', captureStray) // Settlement: exactly one outcome wins; every path funnels through // here, cleans up the timers/listeners, terminates the worker, and @@ -404,7 +388,7 @@ export class WorkerCodeRuntime extends CodeRuntime { // this listener would crash the host process. Junk drops silently. const message = parseWorkerMessage(raw) if (!message) return - if (message.type === 'log' && !settled) admit(message.entry, logs) + if (message.type === 'log' && !settled) admit(message.text, logs) onCall(message) onDone(message) }) diff --git a/packages/code-runtime/code-runtime-worker/src/protocol.ts b/packages/code-runtime/code-runtime-worker/src/protocol.ts index b8ea122c5b..65e6a0d60e 100644 --- a/packages/code-runtime/code-runtime-worker/src/protocol.ts +++ b/packages/code-runtime/code-runtime-worker/src/protocol.ts @@ -9,8 +9,6 @@ * @module @deepseek-ai/dsh-code-runtime-worker/src/protocol */ -import type { CodeLogEntry } from '@deepseek-ai/dsh-code-runtime' - /** What the host hands the worker at spawn, via `workerData`. */ export interface WorkerBootData { /** The type-stripped (plain JS) program body. */ @@ -36,10 +34,10 @@ export interface CallMessage { args: unknown } -/** Worker → host: one captured log entry, streamed eagerly so output survives a mid-run termination (timeout, abort, OOM). */ -export interface LogMessage { +/** Worker → host: captured text, streamed eagerly so output survives a mid-run termination (timeout, abort, OOM). */ +interface LogMessage { type: 'log' - entry: CodeLogEntry + text: string } /** diff --git a/packages/code-runtime/code-runtime-worker/tests/bootstrap.spec.ts b/packages/code-runtime/code-runtime-worker/tests/bootstrap.spec.ts index e41f4455bb..111aa4f15f 100644 --- a/packages/code-runtime/code-runtime-worker/tests/bootstrap.spec.ts +++ b/packages/code-runtime/code-runtime-worker/tests/bootstrap.spec.ts @@ -1,9 +1,8 @@ import { describe, expect, it } from 'vitest' import { EventEmitter } from 'node:events' -import { LogBuffer, makeConsoleShim, makeNamespaces, captureStreamWrites, prepareValue, runWorkerMain, truncateUtf8Bytes, wireReplies } from '@deepseek-ai/dsh-code-runtime-worker/src/bootstrap.ts' -import type { BootstrapPort, PatchableStream, PendingCall } from '@deepseek-ai/dsh-code-runtime-worker/src/bootstrap.ts' -import type { ReplyMessage, WorkerToHost } from '@deepseek-ai/dsh-code-runtime-worker/src/protocol.ts' -import type { CodeLogEntry } from '@deepseek-ai/dsh-code-runtime' +import { LogBuffer, makeConsoleShim, makeNamespaces, captureStreamWrites, prepareValue, runWorkerMain, truncateUtf8Bytes, wireReplies } from '../src/bootstrap.ts' +import type { BootstrapPort, PatchableStream, PendingCall } from '../src/bootstrap.ts' +import type { ReplyMessage, WorkerToHost } from '../src/protocol.ts' /** * An in-process stand-in for the worker's parentPort: the test plays the @@ -31,8 +30,8 @@ class FakePort implements BootstrapPort { this.emitter.emit('message', message) } - logs(): CodeLogEntry[] { - return this.sent.filter(message => message.type === 'log').map(message => message.entry) + logs(): string[] { + return this.sent.filter(message => message.type === 'log').map(message => message.text) } done(): WorkerToHost | undefined { @@ -48,12 +47,12 @@ const BOOT = { maxLogBytes: 65_536, maxValueBytes: 32_768 } describe('LogBuffer', () => { it('streams entries to the sink until the byte budget, then emits one marker and drops the rest', () => { - const seen: CodeLogEntry[] = [] - const buffer = new LogBuffer(10, entry => seen.push(entry)) - buffer.push({ source: 'console', level: 'log', text: '12345' }) - buffer.push({ source: 'console', level: 'log', text: '123456' }) - buffer.push({ source: 'console', level: 'log', text: 'dropped' }) - expect(seen.map(entry => entry.text)).toEqual([ + const seen: string[] = [] + const buffer = new LogBuffer(10, text => seen.push(text)) + buffer.push('12345') + buffer.push('123456') + buffer.push('dropped') + expect(seen).toEqual([ '12345', '[dsh-code-runtime-worker] log capture truncated at 10 bytes', ]) @@ -61,40 +60,37 @@ describe('LogBuffer', () => { }) describe('makeConsoleShim', () => { - it('captures the five levels and renders non-strings inspect-style', () => { - const seen: CodeLogEntry[] = [] - const shim = makeConsoleShim(new LogBuffer(1_000, entry => seen.push(entry))) + it('captures the five methods and renders non-strings inspect-style', () => { + const seen: string[] = [] + const shim = makeConsoleShim(new LogBuffer(1_000, text => seen.push(text))) shim.log('plain', { a: 1 }) shim.info('i') shim.warn('w') shim.error('e') shim.debug('d') - expect(seen.map(entry => entry.level)).toEqual(['log', 'info', 'warn', 'error', 'debug']) - expect(seen[0]?.text).toBe('plain { a: 1 }') - expect(seen.every(entry => entry.source === 'console')).toBe(true) + expect(seen).toEqual(['plain { a: 1 }', 'i', 'w', 'e', 'd']) }) }) describe('captureStreamWrites', () => { it('redirects writes into the buffer and restores on request', () => { - const seen: CodeLogEntry[] = [] - const buffer = new LogBuffer(1_000, entry => seen.push(entry)) + const seen: string[] = [] + const buffer = new LogBuffer(1_000, text => seen.push(text)) let underlying = '' const stream: PatchableStream = { write: (chunk: unknown) => { underlying += String(chunk); return true } } - const restore = captureStreamWrites(buffer, stream, 'stdout') + const restore = captureStreamWrites(buffer, stream) stream.write('captured', 'utf8') stream.write(Buffer.from('bytes')) restore() stream.write('after') - expect(seen.map(entry => entry.text)).toEqual(['captured', 'bytes']) - expect(seen[0]).toMatchObject({ source: 'stdout' }) + expect(seen).toEqual(['captured', 'bytes']) expect(underlying).toBe('after') }) it('invokes the write callback asynchronously, in both optional-encoding shapes', async () => { const buffer = new LogBuffer(1_000, () => {}) const stream: PatchableStream = { write: () => true } - captureStreamWrites(buffer, stream, 'stdout') + captureStreamWrites(buffer, stream) const calls: (Error | null | undefined)[] = [] stream.write('two-arg', (error?: Error | null) => calls.push(error)) stream.write('three-arg', 'utf8', (error?: Error | null) => calls.push(error)) @@ -107,7 +103,7 @@ describe('captureStreamWrites', () => { it('still fires the callback for a write the exhausted budget drops', async () => { const buffer = new LogBuffer(4, () => {}) const stream: PatchableStream = { write: () => true } - captureStreamWrites(buffer, stream, 'stdout') + captureStreamWrites(buffer, stream) stream.write('this write overflows the budget and is dropped') await new Promise(resolve => stream.write('also dropped', resolve)) }) @@ -210,7 +206,7 @@ describe('runWorkerMain', () => { code: 'const doubled = await tools.double({ n: 21 }); console.log("got", doubled); return { doubled };', namespaces: [{ global: 'tools', names: ['double'] }], }, fakeStreams()) - expect(port.logs()).toEqual([{ source: 'console', level: 'log', text: 'got 42' }]) + expect(port.logs()).toEqual(['got 42']) expect(port.done()).toEqual({ type: 'done', value: { doubled: 42 } }) }) @@ -268,6 +264,6 @@ describe('runWorkerMain', () => { // The patch stays installed for the worker's lifetime; writes during the // program landed in order. Here the program wrote nothing via streams, so // only the post-run write above went through the patched slot. - expect(port.logs().at(-1)).toMatchObject({ source: 'stdout' }) + expect(port.logs().at(-1)).toBe('never seen — already restored? no: patch persists in worker') }) }) diff --git a/packages/code-runtime/code-runtime-worker/tests/built-lib.e2e.ts b/packages/code-runtime/code-runtime-worker/tests/built-lib.e2e.ts index aac0ece8a1..2870bf8d4a 100644 --- a/packages/code-runtime/code-runtime-worker/tests/built-lib.e2e.ts +++ b/packages/code-runtime/code-runtime-worker/tests/built-lib.e2e.ts @@ -47,9 +47,9 @@ describe.skipIf(!built)('built lib real load path (plain node)', () => { expect(exitCode, `stderr:\n${stderr}`).toBe(0) const lastLine = stdout.trim().split('\n').at(-1) ?? '' - const result = JSON.parse(lastLine) as { value?: unknown; logs: { source: string; level?: string; text: string }[]; error?: unknown } + const result = JSON.parse(lastLine) as { value?: unknown; logs: string[]; error?: unknown } expect(result.error).toBeUndefined() expect(result.value).toBe(42) - expect(result.logs).toContainEqual({ source: 'console', level: 'log', text: 'halfway 42' }) + expect(result.logs).toContain('halfway 42') }) }) diff --git a/packages/code-runtime/code-runtime-worker/tests/runtime.spec.ts b/packages/code-runtime/code-runtime-worker/tests/runtime.spec.ts index edc2bd1271..9754564702 100644 --- a/packages/code-runtime/code-runtime-worker/tests/runtime.spec.ts +++ b/packages/code-runtime/code-runtime-worker/tests/runtime.spec.ts @@ -28,7 +28,7 @@ describe('WorkerCodeRuntime — programs and bindings (real workers)', () => { expect(runtime.isolation).toBe('worker-thread') }) - it('runs TypeScript (erasable syntax), captures console/stdout in order, returns the value', async () => { + it('runs TypeScript (erasable syntax), captures output in order, returns the value', async () => { const { runtime } = await setup() const result = await runtime.run({ program: ` @@ -43,12 +43,7 @@ describe('WorkerCodeRuntime — programs and bindings (real workers)', () => { }) expect(result.error).toBeUndefined() expect(result.value).toBe(3) - expect(result.logs.map(entry => [entry.source, entry.level ?? null])).toEqual([ - ['console', 'log'], - ['stdout', null], - ['console', 'warn'], - ]) - expect(result.logs[0]?.text).toBe('point { x: 1, y: 2 }') + expect(result.logs).toEqual(['point { x: 1, y: 2 }', 'raw-out\n', 'careful']) }) it('bridges binding calls both ways and rejects the program-side call on a host rejection', async () => { @@ -115,7 +110,7 @@ describe('WorkerCodeRuntime — programs and bindings (real workers)', () => { bindings: [], }) expect(result.error?.kind).toBe('exception') - expect(result.logs.map(entry => entry.text)).toContain('before') + expect(result.logs).toContain('before') }) }) @@ -210,8 +205,8 @@ describe('WorkerCodeRuntime — budgets and containment (real workers)', () => { program: 'for (let i = 0; i < 1000; i++) console.log("spam line", i); return 1', bindings: [], }) - expect(result.logs.at(-1)?.text).toContain('truncated at 300 bytes') - const total = result.logs.reduce((sum, entry) => sum + Buffer.byteLength(entry.text, 'utf8'), 0) + expect(result.logs.at(-1)).toContain('truncated at 300 bytes') + const total = result.logs.reduce((sum, text) => sum + Buffer.byteLength(text, 'utf8'), 0) expect(total).toBeLessThan(1_000) }) @@ -241,7 +236,7 @@ describe('WorkerCodeRuntime — budgets and containment (real workers)', () => { }) expect(result.error).toBeUndefined() expect(result.value).toBe('done') - expect(result.logs).toContainEqual({ source: 'stdout', text: 'flushed' }) + expect(result.logs).toContain('flushed') }) it('caps a huge container whose bounded rendering is small (wire size, not rendering, is what counts)', async () => { @@ -270,8 +265,8 @@ describe('WorkerCodeRuntime — budgets and containment (real workers)', () => { bindings: [], }) expect(result.error).toBeUndefined() - expect(result.logs).toContainEqual({ source: 'stdout', text: 'abcd' }) - expect(result.logs.map(entry => entry.text)).not.toContain('ef') + expect(result.logs).toContain('abcd') + expect(result.logs).not.toContain('ef') }, 15_000) }) @@ -306,11 +301,9 @@ describe('WorkerCodeRuntime — hostile programs (real workers)', () => { { type: 'call', id: 1e9, global: 7, name: 'real', args: {} }, { type: 'call', id: 1e9, global: 'tools', name: 7, args: {} }, { type: 'log' }, - { type: 'log', entry: null }, - { type: 'log', entry: { source: 'stdout', text: 7 } }, - { type: 'log', entry: { source: 'nope', text: 'x' } }, - { type: 'log', entry: { source: 'console', level: 'nope', text: 'x' } }, - { type: 'log', entry: { source: 'console', level: 7, text: 'x' } }, + { type: 'log', text: null }, + { type: 'log', text: 7 }, + { type: 'log', text: {} }, { type: 'done', error: 5 }, { type: 'done', error: { message: 5 } }, ]) parentPort.postMessage(junk); @@ -331,7 +324,7 @@ describe('WorkerCodeRuntime — hostile programs (real workers)', () => { // code and an unbounded result. program: ` const { parentPort } = await import('node:worker_threads'); - for (let i = 0; i < 50; i++) parentPort.postMessage({ type: 'log', entry: { source: 'stdout', text: 'F'.repeat(100), forged: true } }); + for (let i = 0; i < 50; i++) parentPort.postMessage({ type: 'log', text: 'F'.repeat(100), forged: true }); parentPort.postMessage({ type: 'done', value: 'V'.repeat(100000) }); for (;;) {} `, @@ -343,10 +336,9 @@ describe('WorkerCodeRuntime — hostile programs (real workers)', () => { expect(value.endsWith('… [truncated]')).toBe(true) expect(value.length).toBeLessThan(120) const marker = '[dsh-code-runtime-worker] log capture truncated at 200 bytes' - const total = result.logs.reduce((sum, entry) => sum + Buffer.byteLength(entry.text, 'utf8'), 0) + const total = result.logs.reduce((sum, text) => sum + Buffer.byteLength(text, 'utf8'), 0) expect(total).toBeLessThanOrEqual(200 + Buffer.byteLength(marker, 'utf8')) - expect(result.logs.at(-1)?.text).toBe(marker) - expect(result.logs.every(entry => !('forged' in entry))).toBe(true) + expect(result.logs.at(-1)).toBe(marker) }) it('accepts a forged done carrying both value and error (self-sabotage, contained)', async () => { diff --git a/packages/code-runtime/code-runtime/README.md b/packages/code-runtime/code-runtime/README.md index 20c9274b9c..20a92ed526 100644 --- a/packages/code-runtime/code-runtime/README.md +++ b/packages/code-runtime/code-runtime/README.md @@ -16,4 +16,4 @@ Semantics every implementation must honor (contract details in the class JSDoc): ## Vocabulary -`CodeRunRequest` (`program`, `bindings`, `signal?`) carries everything the runtime acts on — defaulting (time budgets, output caps) is the implementation's validated config, never a hidden `??` inside `run()`. `bindings` is a list of `CodeBindingNamespace`s (`global` + `functions`), each exposed to the program as one global object of async callables. `CodeRunResult` reports the completion `value?`, the ordered `logs` (`CodeLogEntry`: `console`/`stdout`/`stderr` source, console `level`, capped text), and the `error?` (`CodeRunFailure`: `kind` + model-feedable `message`). See `src/types.ts` for the full contracts. +`CodeRunRequest` (`program`, `bindings`, `signal?`) carries everything the runtime acts on — defaulting (time budgets, output caps) is the implementation's validated config, never a hidden `??` inside `run()`. `bindings` is a list of `CodeBindingNamespace`s (`global` + `functions`), each exposed to the program as one global object of async callables. `CodeRunResult` reports the completion `value?`, ordered capped `logs: string[]`, and the `error?` (`CodeRunFailure`: `kind` + model-feedable `message`). See `src/types.ts` for the full contracts. diff --git a/packages/code-runtime/code-runtime/src/index.ts b/packages/code-runtime/code-runtime/src/index.ts index 5595469afe..8548b4adb2 100644 --- a/packages/code-runtime/code-runtime/src/index.ts +++ b/packages/code-runtime/code-runtime/src/index.ts @@ -22,7 +22,6 @@ import type { CodeRunRequest, CodeRunResult } from './types.ts' export type { CodeBindingFunction, CodeBindingNamespace, - CodeLogEntry, CodeRunFailure, CodeRunRequest, CodeRunResult, diff --git a/packages/code-runtime/code-runtime/src/types.ts b/packages/code-runtime/code-runtime/src/types.ts index 8278f33a39..d7669a4785 100644 --- a/packages/code-runtime/code-runtime/src/types.ts +++ b/packages/code-runtime/code-runtime/src/types.ts @@ -54,20 +54,6 @@ export interface CodeRunRequest { signal?: AbortSignal } -/** - * One captured output entry, in emission order. `source` says which channel - * produced it: the program's `console` (shimmed by the runtime), or a stray - * write to the underlying stdout/stderr streams. - */ -export interface CodeLogEntry { - /** Which channel produced the text. */ - source: 'console' | 'stdout' | 'stderr' - /** The console method used; present only when `source` is `'console'`. */ - level?: 'log' | 'info' | 'warn' | 'error' | 'debug' - /** The captured text (possibly truncated by the implementation's caps, marked in-band). */ - text: string -} - /** * Why a run failed. The kinds are orthogonal outcomes reported independently * (per docs/defensive-patterns.md): a budget expiry is not an exception, an @@ -98,8 +84,8 @@ export interface CodeRunResult { * or value-less run leaves this absent. */ value?: unknown - /** Everything the program emitted, in order (capped by the implementation). */ - logs: CodeLogEntry[] + /** Text the program emitted, in order (capped by the implementation). */ + logs: string[] /** Present iff the run failed; see {@link CodeRunFailure} for the taxonomy. */ error?: CodeRunFailure } diff --git a/packages/code-runtime/code-runtime/tests/service.spec.ts b/packages/code-runtime/code-runtime/tests/service.spec.ts index 4ff6d8f313..7811ef0531 100644 --- a/packages/code-runtime/code-runtime/tests/service.spec.ts +++ b/packages/code-runtime/code-runtime/tests/service.spec.ts @@ -55,7 +55,7 @@ describe('CodeRuntime service seam', () => { it('reports a failed run as an error field on a resolved result, never a rejection', async () => { const { runtime } = await setup() runtime.nextResult = { - logs: [{ source: 'console', level: 'error', text: 'boom' }], + logs: ['boom'], error: { kind: 'exception', message: 'boom' }, } const result = await runtime.run({ program: 'throw new Error("boom")', bindings: [] }) diff --git a/packages/cordis/tool-cordis/src/api-catalog.ts b/packages/cordis/tool-cordis/src/api-catalog.ts index 51a8f52082..2df7290a3f 100644 --- a/packages/cordis/tool-cordis/src/api-catalog.ts +++ b/packages/cordis/tool-cordis/src/api-catalog.ts @@ -598,10 +598,6 @@ export const TYPE_API: readonly TypeApiEntry[] = [ name: 'CodeBindingNamespace', declaration: 'export interface CodeBindingNamespace {\n global: string;\n functions: Record;\n}', }, - { - name: 'CodeLogEntry', - declaration: 'export interface CodeLogEntry {\n source: \'console\' | \'stdout\' | \'stderr\';\n level?: \'log\' | \'info\' | \'warn\' | \'error\' | \'debug\';\n text: string;\n}', - }, { name: 'CodeRunFailure', declaration: 'export interface CodeRunFailure {\n kind: \'exception\' | \'timeout\' | \'abort\' | \'worker-exit\';\n message: string;\n}', @@ -612,7 +608,7 @@ export const TYPE_API: readonly TypeApiEntry[] = [ }, { name: 'CodeRunResult', - declaration: 'export interface CodeRunResult {\n value?: unknown;\n logs: CodeLogEntry[];\n error?: CodeRunFailure;\n}', + declaration: 'export interface CodeRunResult {\n value?: unknown;\n logs: string[];\n error?: CodeRunFailure;\n}', }, { name: 'CollectedOutput', diff --git a/packages/core/tools/src/code-mode.ts b/packages/core/tools/src/code-mode.ts index 323fbeed2b..8ed73383c1 100644 --- a/packages/core/tools/src/code-mode.ts +++ b/packages/core/tools/src/code-mode.ts @@ -126,14 +126,13 @@ function renderValue(value: unknown): string { /** The run_code result's `meta` payload (JSON-serializable; `presentResult` narrows it back). */ interface RunCodeMeta { logs: CodeRunResult['logs'] - dispatches: number } /** Soft-narrow a result `meta` back to {@link RunCodeMeta} (replay may carry older shapes; presentation must not throw). */ function asRunCodeMeta(meta: unknown): RunCodeMeta | undefined { if (typeof meta !== 'object' || meta === null) return undefined const m = meta as Record - if (!Array.isArray(m.logs) || typeof m.dispatches !== 'number') return undefined + if (!Array.isArray(m.logs) || !m.logs.every(log => typeof log === 'string')) return undefined return m as unknown as RunCodeMeta } @@ -283,12 +282,12 @@ export function createRunCodeTool(registry: ToolRegistry, requireRuntime: () => } if (result.error) { - const logsText = result.logs.length > 0 ? `\nCaptured output:\n${result.logs.map(entry => entry.text).join('\n')}` : '' + const logsText = result.logs.length > 0 ? `\nCaptured output:\n${result.logs.join('\n')}` : '' throw new CodeRunFailedError(`code run failed (${result.error.kind}): ${result.error.message}${logsText}`) } const rendered = renderValue(result.value) - const parts = [result.logs.map(entry => entry.text).join('\n'), rendered].filter(part => part.length > 0) - const meta: RunCodeMeta = { logs: result.logs, dispatches } + const parts = [result.logs.join('\n'), rendered].filter(part => part.length > 0) + const meta: RunCodeMeta = { logs: result.logs } return { content: [{ type: 'text', text: parts.length > 0 ? parts.join('\n') : '(run_code completed with no output)' }], meta, @@ -316,7 +315,7 @@ export function createRunCodeTool(registry: ToolRegistry, requireRuntime: () => presentResult: (_args, result) => { const meta = asRunCodeMeta(result.meta) if (!meta) return undefined - const output = meta.logs.map(entry => entry.text).join('\n') + const output = meta.logs.join('\n') return { card: 'generic', ...output.length > 0 ? { content: [{ type: 'text' as const, text: output }] } : {}, diff --git a/packages/core/tools/tests/code-mode.spec.ts b/packages/core/tools/tests/code-mode.spec.ts index f9d84aadbf..1ae7aae4ec 100644 --- a/packages/core/tools/tests/code-mode.spec.ts +++ b/packages/core/tools/tests/code-mode.spec.ts @@ -328,7 +328,7 @@ describe('the run_code dispatch bridge', () => { const tools = request.bindings[0]!.functions const first = await tools.echo!({ value: 'one' }) const second = await tools.echo!({ value: 'two' }) - return { logs: [{ source: 'console', level: 'log', text: `saw ${String(first)}` }], value: second } + return { logs: [`saw ${String(first)}`], value: second } } const result = await runCode(ctx, 'const …: string = …', { agent }) expect(result.isError).toBe(false) @@ -339,7 +339,7 @@ describe('the run_code dispatch bridge', () => { { parentCallId: 'call-1', subCallId: 'call-1:code:1', name: 'echo', arguments: { value: 'one' }, isError: false, resultSummary: 'echo:one' }, { parentCallId: 'call-1', subCallId: 'call-1:code:2', name: 'echo', arguments: { value: 'two' }, isError: false, resultSummary: 'echo:two' }, ]) - expect(result.meta).toEqual({ logs: [{ source: 'console', level: 'log', text: 'saw echo:one' }], dispatches: 2 }) + expect(result.meta).toEqual({ logs: ['saw echo:one'] }) }) it('exposes only an opaque parent token to nested result observers', async () => { @@ -505,7 +505,7 @@ describe('the run_code dispatch bridge', () => { it('converts a failed run into a structured isError result carrying kind, message, and captured logs', async () => { const { ctx, runtime } = await setup({ mode: 'code' }) runtime.behavior = () => Promise.resolve({ - logs: [{ source: 'console', level: 'log', text: 'got this far' }], + logs: ['got this far'], error: { kind: 'timeout', message: 'compute budget exhausted (300ms busy)' }, }) const result = await runCode(ctx, 'program') @@ -632,7 +632,7 @@ describe('the run_code dispatch bridge', () => { const view = tool.presentResult?.({ code: 'return 1' }, { content: [{ type: 'text', text: 'model-facing' }], isError: false, - meta: { logs: [{ source: 'console', level: 'log', text: 'printed' }], dispatches: 1 }, + meta: { logs: ['printed'] }, }) // The result omits the title — an update replaces only provided fields, // so the pending card's program title persists through completion. @@ -641,9 +641,10 @@ describe('the run_code dispatch bridge', () => { content: [{ type: 'text', text: 'printed' }], }) // No captured output → no content either; everything pending persists. - expect(tool.presentResult?.({ code: 'x' }, { content: [], isError: false, meta: { logs: [], dispatches: 2 } })) + expect(tool.presentResult?.({ code: 'x' }, { content: [], isError: false, meta: { logs: [] } })) .toEqual({ card: 'generic' }) // Replay with an unrecognizable meta falls back to the generic rendering. + expect(tool.presentResult?.({ code: 'x' }, { content: [], isError: false, meta: { logs: [{ text: 'legacy' }], dispatches: 1 } })).toBeUndefined() expect(tool.presentResult?.({ code: 'x' }, { content: [], isError: false, meta: { other: true } })).toBeUndefined() expect(tool.presentResult?.({ code: 'x' }, { content: [], isError: false })).toBeUndefined() }) diff --git a/scripts/type-equiv.manifest.json b/scripts/type-equiv.manifest.json index a219ea1b1d..d3572387aa 100644 --- a/scripts/type-equiv.manifest.json +++ b/scripts/type-equiv.manifest.json @@ -92,7 +92,6 @@ { "doc": "docs/core-data-structures/code-runtime.md", "symbol": "CodeRunResult", "source": "packages/code-runtime/code-runtime/src/types.ts" }, { "doc": "docs/core-data-structures/code-runtime.md", "symbol": "CodeBindingNamespace", "source": "packages/code-runtime/code-runtime/src/types.ts" }, { "doc": "docs/core-data-structures/code-runtime.md", "symbol": "CodeBindingFunction", "source": "packages/code-runtime/code-runtime/src/types.ts" }, - { "doc": "docs/core-data-structures/code-runtime.md", "symbol": "CodeLogEntry", "source": "packages/code-runtime/code-runtime/src/types.ts" }, { "doc": "docs/core-data-structures/code-runtime.md", "symbol": "CodeRunFailure", "source": "packages/code-runtime/code-runtime/src/types.ts" }, { "doc": "docs/core-data-structures/filesystem.md", "symbol": "FsTarget", "source": "packages/fs/fs/src/types.ts" }, From a11396e43cb57fd1523784dcfbe621f68f908967 Mon Sep 17 00:00:00 2001 From: Tianyi Cui <53024+tianyicui@users.noreply.github.com> Date: Tue, 14 Jul 2026 03:09:53 +0800 Subject: [PATCH 22/26] test: refresh code runtime snapshots --- .../acp-agent/tests/snapshots/advanced-toolchain/session.jsonl | 2 +- examples/acp-agent/tests/snapshots/both-mode-turn/session.jsonl | 2 +- examples/acp-agent/tests/snapshots/code-mode-turn/session.jsonl | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) diff --git a/examples/acp-agent/tests/snapshots/advanced-toolchain/session.jsonl b/examples/acp-agent/tests/snapshots/advanced-toolchain/session.jsonl index bb74382b22..ac833c32d8 100644 --- a/examples/acp-agent/tests/snapshots/advanced-toolchain/session.jsonl +++ b/examples/acp-agent/tests/snapshots/advanced-toolchain/session.jsonl @@ -21,7 +21,7 @@ {"type":"assistant/message","seq":19,"time":1783957884490,"data":{"turn":1,"step":2,"content":[{"type":"tool-call","id":"advanced-code","name":"run_code","arguments":"{\"code\":\"return await tools.cordis_inspect({ what: 'dynamic' })\"}"}],"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[14,15,16,17,18],"surfaceOp":"append"} {"type":"tool/call","seq":20,"time":1783957884490,"data":{"turn":1,"step":2,"callId":"advanced-code","name":"run_code","arguments":"{\"code\":\"return await tools.cordis_inspect({ what: 'dynamic' })\"}"}} {"type":"tool/code-dispatch","seq":21,"time":1783957884560,"data":{"parentCallId":"advanced-code","subCallId":"advanced-code:code:1","name":"cordis_inspect","arguments":{"what":"dynamic"},"isError":false,"resultSummary":"## dynamic\n- dyn-1: snapshot-marker [active]"}} -{"type":"tool/result","seq":22,"time":1783957884561,"data":{"turn":1,"step":2,"callId":"advanced-code","content":[{"type":"text","text":"## dynamic\n- dyn-1: snapshot-marker [active]"}],"isError":false,"meta":{"logs":[],"dispatches":1}},"sourceEventSeqs":[20],"surfaceOp":"append"} +{"type":"tool/result","seq":22,"time":1783957884561,"data":{"turn":1,"step":2,"callId":"advanced-code","content":[{"type":"text","text":"## dynamic\n- dyn-1: snapshot-marker [active]"}],"isError":false,"meta":{"logs":[]}},"sourceEventSeqs":[20],"surfaceOp":"append"} {"type":"step/end","seq":23,"time":1783957884561,"data":{"turn":1,"step":2}} {"type":"step/start","seq":24,"time":1783957884562,"data":{"turn":1,"step":3}} {"type":"assistant/chunk","seq":25,"time":1783950000026,"data":{"turn":1,"step":3,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} diff --git a/examples/acp-agent/tests/snapshots/both-mode-turn/session.jsonl b/examples/acp-agent/tests/snapshots/both-mode-turn/session.jsonl index b49097188e..926b1e3389 100644 --- a/examples/acp-agent/tests/snapshots/both-mode-turn/session.jsonl +++ b/examples/acp-agent/tests/snapshots/both-mode-turn/session.jsonl @@ -87,7 +87,7 @@ {"type":"assistant/message","seq":85,"time":1783611775503,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to use the `run_code` tool to execute a program that calls `tools.bash` with the command `echo BOTH_OK` and returns its output."},{"type":"tool-call","id":"call_00_AZFzvUwuC4vAUoICrfke5147","name":"run_code","arguments":"{\"code\": \"const result = await tools.bash({ command: \\\"echo BOTH_OK\\\", description: \\\"Echo BOTH_OK\\\" });\\nreturn result;\"}"}],"usage":{"inputTokens":5530,"outputTokens":108,"cacheReadTokens":0,"reasoningTokens":37}},"sourceEventSeqs":[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84],"surfaceOp":"append"} {"type":"tool/call","seq":86,"time":1783611775504,"data":{"turn":1,"step":1,"callId":"call_00_AZFzvUwuC4vAUoICrfke5147","name":"run_code","arguments":"{\"code\": \"const result = await tools.bash({ command: \\\"echo BOTH_OK\\\", description: \\\"Echo BOTH_OK\\\" });\\nreturn result;\"}"}} {"type":"tool/code-dispatch","seq":87,"time":1783611775590,"data":{"parentCallId":"call_00_AZFzvUwuC4vAUoICrfke5147","subCallId":"call_00_AZFzvUwuC4vAUoICrfke5147:code:1","name":"bash","arguments":{"command":"echo BOTH_OK","description":"Echo BOTH_OK"},"isError":false,"resultSummary":"BOTH_OK\n"}} -{"type":"tool/result","seq":88,"time":1783611775592,"data":{"turn":1,"step":1,"callId":"call_00_AZFzvUwuC4vAUoICrfke5147","content":[{"type":"text","text":"BOTH_OK\n"}],"isError":false,"meta":{"logs":[],"dispatches":1}},"sourceEventSeqs":[86],"surfaceOp":"append"} +{"type":"tool/result","seq":88,"time":1783611775592,"data":{"turn":1,"step":1,"callId":"call_00_AZFzvUwuC4vAUoICrfke5147","content":[{"type":"text","text":"BOTH_OK\n"}],"isError":false,"meta":{"logs":[]}},"sourceEventSeqs":[86],"surfaceOp":"append"} {"type":"step/end","seq":89,"time":1783611775592,"data":{"turn":1,"step":1}} {"type":"step/start","seq":90,"time":1783611775592,"data":{"turn":1,"step":2}} {"type":"assistant/chunk","seq":91,"time":1783611776183,"data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}} diff --git a/examples/acp-agent/tests/snapshots/code-mode-turn/session.jsonl b/examples/acp-agent/tests/snapshots/code-mode-turn/session.jsonl index 9367e2deb0..7b79bd07cb 100644 --- a/examples/acp-agent/tests/snapshots/code-mode-turn/session.jsonl +++ b/examples/acp-agent/tests/snapshots/code-mode-turn/session.jsonl @@ -112,7 +112,7 @@ {"type":"tool/call","seq":110,"time":1783611772840,"data":{"turn":1,"step":1,"callId":"call_00_DRxnM6R1TThfDwcudW0f2050","name":"run_code","arguments":"{\"code\": \"const out1 = await tools.bash({ command: \\\"echo CODE_ONE\\\", description: \\\"First echo\\\" });\\nconst out2 = await tools.bash({ command: \\\"echo CODE_TWO\\\", description: \\\"Second echo\\\" });\\nreturn out1.trim() + \\\"+\\\" + out2.trim();\"}"}} {"type":"tool/code-dispatch","seq":111,"time":1783611772933,"data":{"parentCallId":"call_00_DRxnM6R1TThfDwcudW0f2050","subCallId":"call_00_DRxnM6R1TThfDwcudW0f2050:code:1","name":"bash","arguments":{"command":"echo CODE_ONE","description":"First echo"},"isError":false,"resultSummary":"CODE_ONE\n"}} {"type":"tool/code-dispatch","seq":112,"time":1783611772936,"data":{"parentCallId":"call_00_DRxnM6R1TThfDwcudW0f2050","subCallId":"call_00_DRxnM6R1TThfDwcudW0f2050:code:2","name":"bash","arguments":{"command":"echo CODE_TWO","description":"Second echo"},"isError":false,"resultSummary":"CODE_TWO\n"}} -{"type":"tool/result","seq":113,"time":1783611772937,"data":{"turn":1,"step":1,"callId":"call_00_DRxnM6R1TThfDwcudW0f2050","content":[{"type":"text","text":"CODE_ONE+CODE_TWO"}],"isError":false,"meta":{"logs":[],"dispatches":2}},"sourceEventSeqs":[110],"surfaceOp":"append"} +{"type":"tool/result","seq":113,"time":1783611772937,"data":{"turn":1,"step":1,"callId":"call_00_DRxnM6R1TThfDwcudW0f2050","content":[{"type":"text","text":"CODE_ONE+CODE_TWO"}],"isError":false,"meta":{"logs":[]}},"sourceEventSeqs":[110],"surfaceOp":"append"} {"type":"step/end","seq":114,"time":1783611772938,"data":{"turn":1,"step":1}} {"type":"step/start","seq":115,"time":1783611772938,"data":{"turn":1,"step":2}} {"type":"assistant/chunk","seq":116,"time":1783611773376,"data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}} From 4f40197e169c6f8dadf756ca514b824a9050f58d Mon Sep 17 00:00:00 2001 From: Tianyi Cui <53024+tianyicui@users.noreply.github.com> Date: Tue, 14 Jul 2026 03:18:16 +0800 Subject: [PATCH 23/26] fix: complete code runtime surface pruning --- packages/code-runtime/code-runtime-worker/src/index.ts | 3 --- .../snapshots/python-sdk-single-exe/advanced/result.json | 6 ++---- .../snapshots/python-sdk-single-exe/advanced/session.jsonl | 2 +- 3 files changed, 3 insertions(+), 8 deletions(-) diff --git a/packages/code-runtime/code-runtime-worker/src/index.ts b/packages/code-runtime/code-runtime-worker/src/index.ts index 1f659ad769..96477cf793 100644 --- a/packages/code-runtime/code-runtime-worker/src/index.ts +++ b/packages/code-runtime/code-runtime-worker/src/index.ts @@ -23,9 +23,6 @@ import { prepareValue, truncateUtf8Bytes } from './bootstrap.ts' import { logTruncationMarker } from './protocol.ts' import type { ReplyMessage, WorkerBootData, WorkerToHost } from './protocol.ts' -export type { BootstrapPort, PatchableStream } from './bootstrap.ts' -export type { CallMessage, DoneMessage, ReplyMessage, WorkerBootData, WorkerToHost } from './protocol.ts' - /** Plugin config: every execution cap, changeable from `cordis.yml` (no hardcoded tunables). */ export interface Config { /** diff --git a/scripts/snapshots/python-sdk-single-exe/advanced/result.json b/scripts/snapshots/python-sdk-single-exe/advanced/result.json index 2ecfc76a2e..ca89b2d9a9 100644 --- a/scripts/snapshots/python-sdk-single-exe/advanced/result.json +++ b/scripts/snapshots/python-sdk-single-exe/advanced/result.json @@ -408,8 +408,7 @@ ], "isError": false, "meta": { - "logs": [], - "dispatches": 1 + "logs": [] } }, "sourceEventSeqs": [ @@ -1606,8 +1605,7 @@ ], "isError": false, "meta": { - "logs": [], - "dispatches": 1 + "logs": [] } }, "sourceEventSeqs": [ diff --git a/scripts/snapshots/python-sdk-single-exe/advanced/session.jsonl b/scripts/snapshots/python-sdk-single-exe/advanced/session.jsonl index bb0ee1d4d0..82fb190727 100644 --- a/scripts/snapshots/python-sdk-single-exe/advanced/session.jsonl +++ b/scripts/snapshots/python-sdk-single-exe/advanced/session.jsonl @@ -22,7 +22,7 @@ {"type":"assistant/message","seq":20,"time":0,"data":{"turn":1,"step":2,"content":[{"type":"tool-call","id":"advanced-code","name":"run_code","arguments":"{\"code\": \"return await tools.snapshot_double({ value: 21 })\"}"}],"usage":{"inputTokens":3,"outputTokens":3}},"sourceEventSeqs":[15,16,17,18,19],"surfaceOp":"append"} {"type":"tool/call","seq":21,"time":0,"data":{"turn":1,"step":2,"callId":"advanced-code","name":"run_code","arguments":"{\"code\": \"return await tools.snapshot_double({ value: 21 })\"}"}} {"type":"tool/code-dispatch","seq":22,"time":0,"data":{"parentCallId":"advanced-code","subCallId":"advanced-code:code:1","name":"snapshot_double","arguments":{"value":21},"isError":false,"resultSummary":"42"}} -{"type":"tool/result","seq":23,"time":0,"data":{"turn":1,"step":2,"callId":"advanced-code","content":[{"type":"text","text":"42"}],"isError":false,"meta":{"logs":[],"dispatches":1}},"sourceEventSeqs":[21],"surfaceOp":"append"} +{"type":"tool/result","seq":23,"time":0,"data":{"turn":1,"step":2,"callId":"advanced-code","content":[{"type":"text","text":"42"}],"isError":false,"meta":{"logs":[]}},"sourceEventSeqs":[21],"surfaceOp":"append"} {"type":"step/end","seq":24,"time":0,"data":{"turn":1,"step":2}} {"type":"step/start","seq":25,"time":0,"data":{"turn":1,"step":3}} {"type":"assistant/chunk","seq":26,"time":0,"data":{"turn":1,"step":3,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}} From 5e9131e824a86dc2fcdaf910569b1fbf7d4cc72c Mon Sep 17 00:00:00 2001 From: Tianyi Cui <53024+tianyicui@users.noreply.github.com> Date: Tue, 14 Jul 2026 03:19:33 +0800 Subject: [PATCH 24/26] chore: sync code runtime surface catalog --- docs/config-catalog.md | 2 +- packages/code-runtime/code-runtime-worker/src/protocol.ts | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/docs/config-catalog.md b/docs/config-catalog.md index f785fdf337..d117c13498 100644 --- a/docs/config-catalog.md +++ b/docs/config-catalog.md @@ -212,7 +212,7 @@ export interface Config { } ``` -Source: [`packages/code-runtime/code-runtime-worker/src/index.ts:30`](../packages/code-runtime/code-runtime-worker/src/index.ts) +Source: [`packages/code-runtime/code-runtime-worker/src/index.ts:27`](../packages/code-runtime/code-runtime-worker/src/index.ts) ## `@deepseek-ai/dsh-compact-basic` diff --git a/packages/code-runtime/code-runtime-worker/src/protocol.ts b/packages/code-runtime/code-runtime-worker/src/protocol.ts index 65e6a0d60e..663e407400 100644 --- a/packages/code-runtime/code-runtime-worker/src/protocol.ts +++ b/packages/code-runtime/code-runtime-worker/src/protocol.ts @@ -22,7 +22,7 @@ export interface WorkerBootData { } /** Worker → host: one bridged binding call. */ -export interface CallMessage { +interface CallMessage { type: 'call' /** Worker-issued correlation id; the host answers each id at most once and ignores duplicates. */ id: number From 26e2fe9356bf1c65a8022dac8492be257ebb722a Mon Sep 17 00:00:00 2001 From: Tianyi Cui <53024+tianyicui@users.noreply.github.com> Date: Tue, 14 Jul 2026 04:41:23 +0800 Subject: [PATCH 25/26] refactor: drop assembled section order echo --- docs/config-catalog.md | 2 +- docs/cordis-catalog/services.md | 2 +- packages/bash/tool-bash/tests/tools.spec.ts | 10 ++++++- .../cordis/tool-cordis/src/api-catalog.ts | 2 +- packages/core/system-prompt/src/index.ts | 9 ++----- .../core/system-prompt/tests/scoped.spec.ts | 2 +- .../system-prompt/tests/system-prompt.spec.ts | 26 +++++++++---------- 7 files changed, 28 insertions(+), 25 deletions(-) diff --git a/docs/config-catalog.md b/docs/config-catalog.md index f785fdf337..674970f703 100644 --- a/docs/config-catalog.md +++ b/docs/config-catalog.md @@ -813,7 +813,7 @@ export interface Config { } ``` -Source: [`packages/core/system-prompt/src/index.ts:227`](../packages/core/system-prompt/src/index.ts) +Source: [`packages/core/system-prompt/src/index.ts:223`](../packages/core/system-prompt/src/index.ts) ## `@deepseek-ai/dsh-tool-cordis` diff --git a/docs/cordis-catalog/services.md b/docs/cordis-catalog/services.md index 28d1904ecf..c236baf23e 100644 --- a/docs/cordis-catalog/services.md +++ b/docs/cordis-catalog/services.md @@ -257,7 +257,7 @@ variable(name: string, provider: (context: AssembleContext) => string | undefine async assemble(context: AssembleContext = {}): Promise ``` -Source: [`packages/core/system-prompt/src/index.ts:342`](../../packages/core/system-prompt/src/index.ts) +Source: [`packages/core/system-prompt/src/index.ts:338`](../../packages/core/system-prompt/src/index.ts) ## `ctx.tools` — `ToolRegistry` diff --git a/packages/bash/tool-bash/tests/tools.spec.ts b/packages/bash/tool-bash/tests/tools.spec.ts index 12b4a53958..6041033314 100644 --- a/packages/bash/tool-bash/tests/tools.spec.ts +++ b/packages/bash/tool-bash/tests/tools.spec.ts @@ -293,9 +293,17 @@ describe('bash tool', () => { it('contributes the exit-code habit as its prompt section (guidance the descriptions cannot carry)', async () => { const ctx = await setup() + ctx.systemPrompt.section({ name: 'test:before-bash', order: 104, text: 'before' }) + ctx.systemPrompt.section({ name: 'test:after-bash', order: 106, text: 'after' }) const assembly = await ctx.systemPrompt.assemble() const section = assembly.sections.find(s => s.name === 'tool:bash') - expect(section?.order).toBe(105) + expect(assembly.sections.map(s => s.name)).toEqual([ + 'harness:identity', + 'deployment:persona', + 'test:before-bash', + 'tool:bash', + 'test:after-bash', + ]) expect(section?.text).toContain('[exit code: N]') }) diff --git a/packages/cordis/tool-cordis/src/api-catalog.ts b/packages/cordis/tool-cordis/src/api-catalog.ts index 51a8f52082..9b8f016ca5 100644 --- a/packages/cordis/tool-cordis/src/api-catalog.ts +++ b/packages/cordis/tool-cordis/src/api-catalog.ts @@ -544,7 +544,7 @@ export const TYPE_API: readonly TypeApiEntry[] = [ }, { name: 'AssembledSection', - declaration: 'export interface AssembledSection {\n name: string;\n order: number;\n text: string;\n}', + declaration: 'export interface AssembledSection {\n name: string;\n text: string;\n}', }, { name: 'BashExecRequest', diff --git a/packages/core/system-prompt/src/index.ts b/packages/core/system-prompt/src/index.ts index b886b4c774..ad1fba9f1e 100644 --- a/packages/core/system-prompt/src/index.ts +++ b/packages/core/system-prompt/src/index.ts @@ -103,10 +103,6 @@ export interface PromptSection { export interface AssembledSection { /** The contributing section's unique name. */ name: string - // TODO(assembled-section-order): drop this output field; registry order has - // already sorted the array, and no production renderer/listener reads it. - /** The contributing section's order (sections arrive sorted ascending). */ - order: number /** The resolved (but not yet interpolated) section text. */ text: string } @@ -619,12 +615,11 @@ export class SystemPrompt extends Service { } const assembly: PromptAssembly = { sections: [...sectionByName.values()] + .sort((a, b) => a.order - b.order) .map(section => ({ name: section.name, - order: section.order, text: typeof section.text === 'function' ? section.text(context) : section.text, - })) - .sort((a, b) => a.order - b.order), + })), tools: orderTools(collected, this.toolOrder, knownNames), variables, } diff --git a/packages/core/system-prompt/tests/scoped.spec.ts b/packages/core/system-prompt/tests/scoped.spec.ts index a452bb6797..aac3f79e88 100644 --- a/packages/core/system-prompt/tests/scoped.spec.ts +++ b/packages/core/system-prompt/tests/scoped.spec.ts @@ -139,7 +139,7 @@ describe('scoped assemble dispatch', () => { scope.ctx.on('system-prompt/assemble', async (_assembly: PromptAssembly, context, next: () => Promise) => { shaped.push(context.scope) const result = await next() - result.sections.push({ name: 'listener:extra', order: 999, text: 'listener text' }) + result.sections.push({ name: 'listener:extra', text: 'listener text' }) return result }) diff --git a/packages/core/system-prompt/tests/system-prompt.spec.ts b/packages/core/system-prompt/tests/system-prompt.spec.ts index c7436711fa..b084104d22 100644 --- a/packages/core/system-prompt/tests/system-prompt.spec.ts +++ b/packages/core/system-prompt/tests/system-prompt.spec.ts @@ -21,9 +21,9 @@ describe('SystemPrompt', () => { await ctx.plugin(SystemPrompt, { persona: 'You are DeepSeek Harness SDK.' }) const assembly = await ctx.systemPrompt.assemble() - expect(assembly.sections.map(s => [s.name, s.order])).toEqual([ - ['harness:identity', -100], - ['deployment:persona', 0], + expect(assembly.sections.map(s => s.name)).toEqual([ + 'harness:identity', + 'deployment:persona', ]) expect(renderPrompt(assembly)).toBe(`${IDENTITY}\n\nYou are DeepSeek Harness SDK.`) // The names are reserved by the plugin — one owner per section. @@ -183,7 +183,7 @@ describe('SystemPrompt', () => { const contexts: AssembleContext[] = [] ctx.on('system-prompt/assemble', async (assembly: PromptAssembly, context, next) => { contexts.push(context) - assembly.sections.push({ name: 'from-a', order: 100, text: 'a' }) + assembly.sections.push({ name: 'from-a', text: 'a' }) return next() }) // Listener B (registered later, runs after A) sees A's contribution. @@ -235,8 +235,8 @@ describe('SystemPrompt', () => { it('filters out empty section text from renderPrompt', () => { const result = renderPrompt({ sections: [ - { name: 'empty', order: 0, text: '' }, - { name: 'real', order: 1, text: 'content' }, + { name: 'empty', text: '' }, + { name: 'real', text: 'content' }, ], tools: [], variables: {}, @@ -356,13 +356,13 @@ describe('SystemPrompt', () => { }) it('names "(none)" when no variables are registered at all', () => { - expect(() => renderPrompt({ sections: [{ name: 's', order: 0, text: '{{x}}' }], tools: [], variables: {} })) + expect(() => renderPrompt({ sections: [{ name: 's', text: '{{x}}' }], tools: [], variables: {} })) .toThrow('unknown prompt variable "{{x}}" in section "s"; registered variables: (none)') }) it('throws when a referenced variable has no value for this assembly', () => { expect(() => renderPrompt({ - sections: [{ name: 'persona', order: 0, text: 'in {{cwd}}' }], + sections: [{ name: 'persona', text: 'in {{cwd}}' }], tools: [], variables: { cwd: undefined }, })).toThrow('prompt variable "{{cwd}}" has no value for this assembly (section "persona")') @@ -370,7 +370,7 @@ describe('SystemPrompt', () => { it('throws on a malformed complete reference, e.g. inner spaces', () => { expect(() => renderPrompt({ - sections: [{ name: 's', order: 0, text: 'on {{ model }}' }], + sections: [{ name: 's', text: 'on {{ model }}' }], tools: [], variables: { model: 'm' }, })).toThrow('malformed prompt variable reference "{{ model }}" in section "s"') @@ -378,7 +378,7 @@ describe('SystemPrompt', () => { it('leaves a lone {{ verbatim only when NO }} follows anywhere after it', () => { const text = renderPrompt({ - sections: [{ name: 's', order: 0, text: 'shell ${X:-{{fallback} stays' }], + sections: [{ name: 's', text: 'shell ${X:-{{fallback} stays' }], tools: [], variables: {}, }) @@ -390,7 +390,7 @@ describe('SystemPrompt', () => { { text: 'x {{a{b}} y {{model}}', label: 'nested brace inside a would-be group' }, ])('throws on a mangled reference with a }} still following ($label)', ({ text }) => { expect(() => renderPrompt({ - sections: [{ name: 's', order: 0, text }], + sections: [{ name: 's', text }], tools: [], variables: { model: 'm' }, })).toThrow('malformed prompt variable reference at') @@ -400,7 +400,7 @@ describe('SystemPrompt', () => { // `in` would find Object.prototype.constructor and splice function // source into the prompt; Object.hasOwn must reject it instead. expect(() => renderPrompt({ - sections: [{ name: 's', order: 0, text: 'on {{constructor}}' }], + sections: [{ name: 's', text: 'on {{constructor}}' }], tools: [], variables: { model: 'm' }, })).toThrow('unknown prompt variable "{{constructor}}"') @@ -416,7 +416,7 @@ describe('SystemPrompt', () => { it('never re-scans substituted values (a value containing {{sneaky}} stays literal)', () => { const text = renderPrompt({ - sections: [{ name: 's', order: 0, text: 'v = {{model}}!' }], + sections: [{ name: 's', text: 'v = {{model}}!' }], tools: [], variables: { model: 'literal {{sneaky}} inside' }, }) From b9eeb2162c2eee306977893772098d608d9c7a43 Mon Sep 17 00:00:00 2001 From: lintianle Date: Wed, 15 Jul 2026 15:40:02 +0800 Subject: [PATCH 26/26] Fix lockfile broken by hand-resolved merge conflict The web-editor conflict resolution renamed the cordis snapshot keys to rc.7 but left the mcp-client importer pointing at the deleted cordis@4.0.0-rc.6(...rc.4) key, so every CI lane failed at 'pnpm install --frozen-lockfile' with ERR_PNPM_LOCKFILE_MISSING_DEPENDENCY. Align the mcp-client cordis range with the repo-wide ^4.0.0-rc.7 sweep from master and re-resolve the lockfile with pnpm. --- packages/mcp/mcp-client/package.json | 4 ++-- pnpm-lock.yaml | 4 ++-- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/packages/mcp/mcp-client/package.json b/packages/mcp/mcp-client/package.json index 638777017d..6b8f145108 100644 --- a/packages/mcp/mcp-client/package.json +++ b/packages/mcp/mcp-client/package.json @@ -24,7 +24,7 @@ "peerDependencies": { "@deepseek-ai/dsh-tools": "^0.0.1", "@deepseek-ai/dsh-llm": "^0.0.1", - "cordis": "^4.0.0-rc.6" + "cordis": "^4.0.0-rc.7" }, "dependencies": { "@modelcontextprotocol/sdk": "^1.12.0", @@ -35,7 +35,7 @@ "@deepseek-ai/dsh-llm": "workspace:^", "@modelcontextprotocol/server-everything": "^2026.7.4", "@modelcontextprotocol/server-filesystem": "^2026.7.4", - "cordis": "^4.0.0-rc.6", + "cordis": "^4.0.0-rc.7", "zod": "^4.4.3" } } diff --git a/pnpm-lock.yaml b/pnpm-lock.yaml index b1b56db805..4e0a2af98a 100644 --- a/pnpm-lock.yaml +++ b/pnpm-lock.yaml @@ -746,8 +746,8 @@ importers: specifier: ^2026.7.4 version: 2026.7.10(zod@4.4.3) cordis: - specifier: ^4.0.0-rc.6 - version: 4.0.0-rc.6(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.4) + specifier: ^4.0.0-rc.7 + version: 4.0.0-rc.7(@cordisjs/plugin-include@1.0.4)(@cordisjs/plugin-loader@1.0.0-rc.5) zod: specifier: ^4.4.3 version: 4.4.3