From 218c98c2c232eee415043cc62009ebca63d12195 Mon Sep 17 00:00:00 2001 From: Turtle Date: Fri, 24 Jul 2026 12:12:18 +0800 Subject: [PATCH 1/8] fix(goal): tolerate strict provider filler fields --- ...26-07-19-model-facing-goal-tools.i18n.yaml | 4 +- .../2026-07-19-model-facing-goal-tools.md | 4 +- .../2026-07-19-model-facing-goal-tools.zh.md | 4 +- .../tool-schemas.expected.json | 8 +-- .../both-mode-turn/tool-schemas.expected.json | 8 +-- .../escalation-approved/session.jsonl | 4 +- .../escalation-rejected/session.jsonl | 4 +- .../tests/snapshots/fs-edit/session.jsonl | 2 +- .../snapshots/fs-policy-reject/session.jsonl | 4 +- .../fs-write-overwrite/session.jsonl | 2 +- .../hook-cc-pretool-ask/session.jsonl | 4 +- .../lsp-definition/tool-schemas.expected.json | 8 +-- .../tool-schemas.expected.json | 16 ++--- .../tool-schemas.expected.json | 16 ++--- .../plan-mode/tool-schemas.expected.json | 16 ++--- .../pty-tools/tool-schemas.expected.json | 8 +-- .../skill-load/tool-schemas.expected.json | 8 +-- .../text-turn/tool-schemas.expected.json | 8 +-- .../tool-schemas.expected.json | 8 +-- packages/goal/tool-goal/README.md | 2 +- packages/goal/tool-goal/src/index.ts | 29 ++------- .../goal/tool-goal/tests/tool-goal.spec.ts | 61 +++++++++++-------- 22 files changed, 109 insertions(+), 119 deletions(-) diff --git a/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.i18n.yaml b/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.i18n.yaml index e53c591aa5..b91a8758ca 100644 --- a/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.i18n.yaml +++ b/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.i18n.yaml @@ -2,5 +2,5 @@ # side as of the last confirmed-consistent state. Both languages carry equal authority; # after editing either side, bring the other along and re-record with: # pnpm run verify-translation-pairing --write -2026-07-19-model-facing-goal-tools.md: 7cc3907d708115207e166455ea988120a03d768b -2026-07-19-model-facing-goal-tools.zh.md: 1a381160354d6a2a24f957f41bc9e375c1ab01ca +2026-07-19-model-facing-goal-tools.md: 7df04678ab8f78be71504b9a0d6e3a05391833e3 +2026-07-19-model-facing-goal-tools.zh.md: 102337fd57291102ce16a3978bafe9fcf62b9695 diff --git a/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.md b/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.md index 7cc3907d70..7df04678ab 100644 --- a/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.md +++ b/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.md @@ -16,7 +16,7 @@ The surface also needs to preserve the separation between durable state and live ### Tools and model contract -`get_goal()` returns the current goal or `null`. A non-null result contains the compare-and-set id and revision, objective, durable phase, admitted and maximum goal rounds, any blocker reason, plus the process-local activation observation. `create_goal(objective, max_goal_rounds?)` creates one long-running same-session objective. `update_goal(goal_id, revision, action, objective?, max_goal_rounds?, blocked_reason?)` supports `edit`, `pause`, `resume`, `complete`, and `blocked`; replacement fields are valid only for `edit`, while a non-empty `blocked_reason` is required only for `blocked` and persists under the stable `model-reported` code. +`get_goal()` returns the current goal or `null`. A non-null result contains the compare-and-set id and revision, objective, durable phase, admitted and maximum goal rounds, any blocker reason, plus the process-local activation observation. `create_goal(objective, max_goal_rounds?)` creates one long-running same-session objective. `update_goal(goal_id, revision, action, objective?, max_goal_rounds?, blocked_reason?)` supports `edit`, `pause`, `resume`, `complete`, and `blocked`; each action consumes only its fields and ignores the others, while `blocked` still requires a non-empty `blocked_reason` and persists it under the stable `model-reported` code. This tolerance is necessary because strict-schema providers may populate every declared property even though JSON Schema marks the action-specific fields optional. The prompt tells the model that it may infer goal intent from a direct human request in any wording or language, but should not convert routine single-turn work into a goal. It must read the current goal before updating and copy the exact id and revision. On a restored or forked active-but-disarmed goal, a semantic human request to continue is grounds for `resume`. Completion is reserved for an achieved objective, and difficulty or uncertainty alone is not a blocker; a block report must name the concrete condition. @@ -38,7 +38,7 @@ Complete and blocked accept either direct-human authority or the exact current g ## Testing -Unit coverage pins registration and disposal, exclusive scheduling, generated prompt policy, generic presentation, direct-human creation in a non-English turn, exact/stale/non-running agent and driver checks, live-child rejection, resumed-fork root authority, steering, mismatched initiators, read/create/edit/pause/resume behavior, conditional blocker explanations, rearming after a session-start edge, authority-before-conditional-argument failures, exact goal-round completion, autonomous-only terminal stopping, the configured blocking threshold, and immediate human blocking. A keyless replay snapshot mounts the goal domain and tools into the real headless one-shot application, drives `create_goal` and `get_goal` through the shipped loop and persistence stack, pins its stream-json transcript, and inspects the externally persisted goal change. The echo-agent fixture is intentionally not used as an application-UX surrogate. +Unit coverage pins registration and disposal, exclusive scheduling, generated prompt policy, generic presentation, direct-human creation in a non-English turn, exact/stale/non-running agent and driver checks, live-child rejection, resumed-fork root authority, steering, mismatched initiators, read/create/edit/pause/resume behavior, required blocker explanations, strict-provider filler fields, rearming after a session-start edge, authority-before-argument failures, exact goal-round completion, autonomous-only terminal stopping, the configured blocking threshold, and immediate human blocking. A keyless replay snapshot mounts the goal domain and tools into the real headless one-shot application, drives `create_goal` and `get_goal` through the shipped loop and persistence stack, pins its stream-json transcript, and inspects the externally persisted goal change. The echo-agent fixture is intentionally not used as an application-UX surrogate. ## Alternatives considered diff --git a/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.zh.md b/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.zh.md index 1a38116035..102337fd57 100644 --- a/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.zh.md +++ b/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.zh.md @@ -16,7 +16,7 @@ Status: implemented ### 工具与模型契约 -`get_goal()` 返回当前目标或 `null`。非空结果包含用于比较并交换的 id 与修订号、目标描述、持久阶段、已接纳和最大目标回合数、可能存在的阻塞原因，以及进程本地激活态观察。`create_goal(objective, max_goal_rounds?)` 创建一个长时间运行的同会话目标。`update_goal(goal_id, revision, action, objective?, max_goal_rounds?, blocked_reason?)` 支持 `edit`、`pause`、`resume`、`complete` 和 `blocked`；替换字段仅对 `edit` 有效，非空的 `blocked_reason` 仅在 `blocked` 时必填，并以稳定代码 `model-reported` 持久化。 +`get_goal()` 返回当前目标或 `null`。非空结果包含用于比较并交换的 id 与修订号、目标描述、持久阶段、已接纳和最大目标回合数、可能存在的阻塞原因，以及进程本地激活态观察。`create_goal(objective, max_goal_rounds?)` 创建一个长时间运行的同会话目标。`update_goal(goal_id, revision, action, objective?, max_goal_rounds?, blocked_reason?)` 支持 `edit`、`pause`、`resume`、`complete` 和 `blocked`；每个动作只消费自己的字段并忽略其他字段，而 `blocked` 仍要求非空的 `blocked_reason`，并以稳定代码 `model-reported` 持久化。尽管 JSON Schema 将动作专用字段标记为可选，严格模式提供方仍可能填充所有已声明属性，因此需要这种容忍行为。 提示词告诉模型：它可以从任何措辞或语言的直接人类请求中推断目标意图，但不应把常规单轮工作转换为目标。更新前必须读取当前目标，并复制准确的 id 和修订号。对于恢复或派生后处于活跃但未激活状态的目标，人类在语义上要求继续即可成为执行 `resume` 的依据。只有目标已经实现时才能标记完成，困难或不确定性本身不构成阻塞；阻塞报告必须说明具体条件。 @@ -38,7 +38,7 @@ Status: implemented ## 测试 -单元测试固定注册与释放、独占调度、生成的提示词策略、通用展示、非英语轮次中的直接人类创建、精确/陈旧/非运行中智能体与驱动检查、实时子智能体拒绝、恢复后派生根的权限、steering、发起者不匹配、读取/创建/编辑/暂停/恢复行为、条件式阻塞说明、会话启动边沿后的重新激活、权限先于条件参数失败、准确目标回合的完成、仅自主回合触发终止、可配置阻塞阈值，以及人类立即阻塞。无密钥回放快照把目标领域和工具挂载到真实的 headless 单次运行应用中，通过随附循环与持久化栈驱动 `create_goal` 和 `get_goal`，固定 stream-json 转录，并检查外部持久化的目标变更。这里有意不把 echo-agent 测试夹具当作应用 UX 的替代品。 +单元测试固定注册与释放、独占调度、生成的提示词策略、通用展示、非英语轮次中的直接人类创建、精确/陈旧/非运行中智能体与驱动检查、实时子智能体拒绝、恢复后派生根的权限、steering、发起者不匹配、读取/创建/编辑/暂停/恢复行为、必需的阻塞说明、严格模式提供方填充字段、会话启动边沿后的重新激活、权限先于参数失败、准确目标回合的完成、仅自主回合触发终止、可配置阻塞阈值，以及人类立即阻塞。无密钥回放快照把目标领域和工具挂载到真实的 headless 单次运行应用中，通过随附循环与持久化栈驱动 `create_goal` 和 `get_goal`，固定 stream-json 转录，并检查外部持久化的目标变更。这里有意不把 echo-agent 测试夹具当作应用 UX 的替代品。 ## 考虑过的替代方案 diff --git a/examples/acp-agent/tests/snapshots/advanced-toolchain/tool-schemas.expected.json b/examples/acp-agent/tests/snapshots/advanced-toolchain/tool-schemas.expected.json index 6b50a5d220..b3f147fdbf 100644 --- a/examples/acp-agent/tests/snapshots/advanced-toolchain/tool-schemas.expected.json +++ b/examples/acp-agent/tests/snapshots/advanced-toolchain/tool-schemas.expected.json @@ -470,7 +470,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -495,15 +495,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective; valid only with action edit." + "description": "Replacement objective for action edit; ignored for other actions." }, "max_goal_rounds": { "type": "number", - "description": "Replacement cap; valid only with action edit." + "description": "Replacement round cap for action edit; ignored for other actions." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition; required only with action blocked." + "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." } }, "required": [ diff --git a/examples/acp-agent/tests/snapshots/both-mode-turn/tool-schemas.expected.json b/examples/acp-agent/tests/snapshots/both-mode-turn/tool-schemas.expected.json index 7ceeec4042..0e2ebef051 100644 --- a/examples/acp-agent/tests/snapshots/both-mode-turn/tool-schemas.expected.json +++ b/examples/acp-agent/tests/snapshots/both-mode-turn/tool-schemas.expected.json @@ -413,7 +413,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -438,15 +438,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective; valid only with action edit." + "description": "Replacement objective for action edit; ignored for other actions." }, "max_goal_rounds": { "type": "number", - "description": "Replacement cap; valid only with action edit." + "description": "Replacement round cap for action edit; ignored for other actions." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition; required only with action blocked." + "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." } }, "required": [ diff --git a/examples/acp-agent/tests/snapshots/escalation-approved/session.jsonl b/examples/acp-agent/tests/snapshots/escalation-approved/session.jsonl index 4438550b97..6aa7fda0a8 100644 --- a/examples/acp-agent/tests/snapshots/escalation-approved/session.jsonl +++ b/examples/acp-agent/tests/snapshots/escalation-approved/session.jsonl @@ -132,8 +132,8 @@ {"type":"assistant/chunk","seq":130,"time":1783962245385,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":131,"time":1783962245385,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run a command with sandbox_permissions set to danger-full-access, no prior run needed, justified as instructed."},{"type":"tool-call","id":"call_00_d0sAHpJ9mYOJi0z7KNy30441","name":"bash","arguments":"{\"command\": \"printf 'escalated\\\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt\", \"description\": \"Write file outside workspace and verify\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to write a file outside the workspace\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":1501,"outputTokens":174,"cacheReadTokens":0,"reasoningTokens":28}},"sourceEventSeqs":[8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130],"surfaceOp":"append"} {"type":"tool/call","seq":132,"time":1783962245385,"data":{"turn":1,"step":1,"callId":"call_00_d0sAHpJ9mYOJi0z7KNy30441","name":"bash","arguments":"{\"command\": \"printf 'escalated\\\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt\", \"description\": \"Write file outside workspace and verify\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to write a file outside the workspace\"}"}} -{"type":"approval/asked","seq":133,"time":1783962245386,"data":{"id":"efa76e4a-8e3e-4be0-9608-a68def7354b9","toolName":"bash","callId":"call_00_d0sAHpJ9mYOJi0z7KNy30441","reason":"escalate sandbox to danger-full-access: the user asked to write a file outside the workspace"}} -{"type":"approval/decided","seq":134,"time":1783962245387,"data":{"id":"efa76e4a-8e3e-4be0-9608-a68def7354b9","outcome":"allowed-once"}} +{"type":"approval/asked","seq":133,"time":1783962245386,"data":{"id":"21634890-d003-489f-b1b5-d38e187561c5","toolName":"bash","callId":"call_00_d0sAHpJ9mYOJi0z7KNy30441","reason":"escalate sandbox to danger-full-access: the user asked to write a file outside the workspace"}} +{"type":"approval/decided","seq":134,"time":1783962245387,"data":{"id":"21634890-d003-489f-b1b5-d38e187561c5","outcome":"allowed-once"}} {"type":"tool/result","seq":135,"time":1783962245399,"data":{"turn":1,"step":1,"callId":"call_00_d0sAHpJ9mYOJi0z7KNy30441","content":[{"type":"text","text":"escalated\n"}],"isError":false},"sourceEventSeqs":[132],"surfaceOp":"append"} {"type":"step/end","seq":136,"time":1783962245400,"data":{"turn":1,"step":1}} {"type":"step/start","seq":137,"time":1783962245400,"data":{"turn":1,"step":2}} diff --git a/examples/acp-agent/tests/snapshots/escalation-rejected/session.jsonl b/examples/acp-agent/tests/snapshots/escalation-rejected/session.jsonl index eb01e7443e..0d384a55ee 100644 --- a/examples/acp-agent/tests/snapshots/escalation-rejected/session.jsonl +++ b/examples/acp-agent/tests/snapshots/escalation-rejected/session.jsonl @@ -156,8 +156,8 @@ {"type":"assistant/chunk","seq":154,"time":1783962246274,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":155,"time":1783962246274,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run a specific command with `sandbox_permissions` set to `danger-full-access` and a specific justification. They explicitly said NOT to run it without sandbox_permissions first. Let me do exactly that."},{"type":"tool-call","id":"call_00_WB1vnPomi8yr6MlcFKTj7912","name":"bash","arguments":"{\"command\": \"printf 'escalated\\\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt\", \"description\": \"Write to /tmp and verify, then clean up\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to write a file outside the workspace\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":1509,"outputTokens":198,"cacheReadTokens":0,"reasoningTokens":48}},"sourceEventSeqs":[8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154],"surfaceOp":"append"} {"type":"tool/call","seq":156,"time":1783962246274,"data":{"turn":1,"step":1,"callId":"call_00_WB1vnPomi8yr6MlcFKTj7912","name":"bash","arguments":"{\"command\": \"printf 'escalated\\\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt\", \"description\": \"Write to /tmp and verify, then clean up\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to write a file outside the workspace\"}"}} -{"type":"approval/asked","seq":157,"time":1783962246275,"data":{"id":"e5cc84ef-8704-4cd9-8e47-897a0a51beb6","toolName":"bash","callId":"call_00_WB1vnPomi8yr6MlcFKTj7912","reason":"escalate sandbox to danger-full-access: the user asked to write a file outside the workspace"}} -{"type":"approval/decided","seq":158,"time":1783962246275,"data":{"id":"e5cc84ef-8704-4cd9-8e47-897a0a51beb6","outcome":"rejected"}} +{"type":"approval/asked","seq":157,"time":1783962246275,"data":{"id":"4ffb95e3-8f94-41db-bfd3-2f56e6631ef8","toolName":"bash","callId":"call_00_WB1vnPomi8yr6MlcFKTj7912","reason":"escalate sandbox to danger-full-access: the user asked to write a file outside the workspace"}} +{"type":"approval/decided","seq":158,"time":1783962246275,"data":{"id":"4ffb95e3-8f94-41db-bfd3-2f56e6631ef8","outcome":"rejected"}} {"type":"tool/result","seq":159,"time":1783962246275,"data":{"turn":1,"step":1,"callId":"call_00_WB1vnPomi8yr6MlcFKTj7912","content":[{"type":"text","text":"Error: the user rejected escalating this command to \"danger-full-access\""}],"isError":true},"sourceEventSeqs":[156],"surfaceOp":"append"} {"type":"step/end","seq":160,"time":1783962246276,"data":{"turn":1,"step":1}} {"type":"step/start","seq":161,"time":1783962246276,"data":{"turn":1,"step":2}} diff --git a/examples/acp-agent/tests/snapshots/fs-edit/session.jsonl b/examples/acp-agent/tests/snapshots/fs-edit/session.jsonl index d0bd2d5bef..bcdd4c7b99 100644 --- a/examples/acp-agent/tests/snapshots/fs-edit/session.jsonl +++ b/examples/acp-agent/tests/snapshots/fs-edit/session.jsonl @@ -130,7 +130,7 @@ {"type":"assistant/chunk","seq":128,"time":1783352087469,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":129,"time":1783352087469,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"Now I need to replace \"DEBUG\" with \"RELEASE\" using the edit tool."},{"type":"tool-call","id":"call_00_vOytneZ0XpsLslEEJAxR6398","name":"edit","arguments":"{\"file_path\": \"config.txt\", \"old_string\": \"DEBUG\", \"new_string\": \"RELEASE\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":241,"outputTokens":98,"cacheReadTokens":2816,"reasoningTokens":18}},"sourceEventSeqs":[74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128],"surfaceOp":"append"} {"type":"tool/call","seq":130,"time":1783352087469,"data":{"turn":1,"step":2,"callId":"call_00_vOytneZ0XpsLslEEJAxR6398","name":"edit","arguments":"{\"file_path\": \"config.txt\", \"old_string\": \"DEBUG\", \"new_string\": \"RELEASE\"}"}} -{"type":"tool/result","seq":131,"time":1783352087476,"data":{"turn":1,"step":2,"callId":"call_00_vOytneZ0XpsLslEEJAxR6398","content":[{"type":"text","text":"The file /tmp/acp-snap-cwd-0BxHdV/config.txt has been updated successfully."}],"isError":false,"meta":{"diffs":[{"path":"config.txt","oldText":"mode=DEBUG\nlevel=info","newText":"mode=RELEASE\nlevel=info"}]}},"sourceEventSeqs":[130],"surfaceOp":"append"} +{"type":"tool/result","seq":131,"time":1783352087476,"data":{"turn":1,"step":2,"callId":"call_00_vOytneZ0XpsLslEEJAxR6398","content":[{"type":"text","text":"The file /private/tmp/acp-snap-cwd-0BxHdV/config.txt has been updated successfully."}],"isError":false,"meta":{"diffs":[{"path":"config.txt","oldText":"mode=DEBUG\nlevel=info","newText":"mode=RELEASE\nlevel=info"}]}},"sourceEventSeqs":[130],"surfaceOp":"append"} {"type":"step/end","seq":132,"time":1783352087477,"data":{"turn":1,"step":2}} {"type":"step/start","seq":133,"time":1783352087477,"data":{"turn":1,"step":3}} {"type":"assistant/chunk","seq":134,"time":1783352088286,"data":{"turn":1,"step":3,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}} diff --git a/examples/acp-agent/tests/snapshots/fs-policy-reject/session.jsonl b/examples/acp-agent/tests/snapshots/fs-policy-reject/session.jsonl index 8f97feff04..25c57ee964 100644 --- a/examples/acp-agent/tests/snapshots/fs-policy-reject/session.jsonl +++ b/examples/acp-agent/tests/snapshots/fs-policy-reject/session.jsonl @@ -78,7 +78,7 @@ {"type":"assistant/chunk","seq":76,"time":1783611703969,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":77,"time":1783611703972,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to use the edit tool to replace \"blue\" with \"green\" in settings.txt without reading the file first, and then reply with just \"DONE\"."},{"type":"tool-call","id":"call_00_x0zlnXl5JOxLrAYL9y7P0119","name":"edit","arguments":"{\"file_path\": \"settings.txt\", \"old_string\": \"blue\", \"new_string\": \"green\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":3132,"outputTokens":115,"cacheReadTokens":0,"reasoningTokens":36}},"sourceEventSeqs":[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],"surfaceOp":"append"} {"type":"tool/call","seq":78,"time":1783611703972,"data":{"turn":1,"step":1,"callId":"call_00_x0zlnXl5JOxLrAYL9y7P0119","name":"edit","arguments":"{\"file_path\": \"settings.txt\", \"old_string\": \"blue\", \"new_string\": \"green\"}"}} -{"type":"tool/result","seq":79,"time":1783611703978,"data":{"turn":1,"step":1,"callId":"call_00_x0zlnXl5JOxLrAYL9y7P0119","content":[{"type":"text","text":"Error: edit requires reading \"/var/folders/2c/psb0_fmx7hbgz558xjt_f0l00000gn/T/acp-snap-cwd-QzoqnB/settings.txt\" first"}],"isError":true,"error":{"name":"FsError","code":"FS_NOT_OBSERVED"}},"sourceEventSeqs":[78],"surfaceOp":"append"} +{"type":"tool/result","seq":79,"time":1783611703978,"data":{"turn":1,"step":1,"callId":"call_00_x0zlnXl5JOxLrAYL9y7P0119","content":[{"type":"text","text":"Error: edit requires reading \"/private/var/folders/2c/psb0_fmx7hbgz558xjt_f0l00000gn/T/acp-snap-cwd-QzoqnB/settings.txt\" first"}],"isError":true,"error":{"name":"FsError","code":"FS_NOT_OBSERVED"}},"sourceEventSeqs":[78],"surfaceOp":"append"} {"type":"step/end","seq":80,"time":1783611703978,"data":{"turn":1,"step":1}} {"type":"step/start","seq":81,"time":1783611703978,"data":{"turn":1,"step":2}} {"type":"assistant/chunk","seq":82,"time":1783611704825,"data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}} @@ -226,7 +226,7 @@ {"type":"assistant/chunk","seq":224,"time":1783611707096,"data":{"turn":1,"step":3,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":225,"time":1783611707097,"data":{"turn":1,"step":3,"content":[{"type":"reasoning","text":"The file contains \"color: blue\". I need to replace \"blue\" with \"green\". The edit tool said it requires reading first — now I've read it, so the edit should work."},{"type":"tool-call","id":"call_00_GVknJu2tksKkP4lALCwh0926","name":"edit","arguments":"{\"file_path\": \"settings.txt\", \"old_string\": \"blue\", \"new_string\": \"green\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":281,"outputTokens":119,"cacheReadTokens":3200,"reasoningTokens":40}},"sourceEventSeqs":[149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188,189,190,191,192,193,194,195,196,197,198,199,200,201,202,203,204,205,206,207,208,209,210,211,212,213,214,215,216,217,218,219,220,221,222,223,224],"surfaceOp":"append"} {"type":"tool/call","seq":226,"time":1783611707097,"data":{"turn":1,"step":3,"callId":"call_00_GVknJu2tksKkP4lALCwh0926","name":"edit","arguments":"{\"file_path\": \"settings.txt\", \"old_string\": \"blue\", \"new_string\": \"green\"}"}} -{"type":"tool/result","seq":227,"time":1783611707114,"data":{"turn":1,"step":3,"callId":"call_00_GVknJu2tksKkP4lALCwh0926","content":[{"type":"text","text":"The file /var/folders/2c/psb0_fmx7hbgz558xjt_f0l00000gn/T/acp-snap-cwd-QzoqnB/settings.txt has been updated successfully."}],"isError":false,"meta":{"diffs":[{"path":"settings.txt","oldText":"color: blue","newText":"color: green"}]}},"sourceEventSeqs":[226],"surfaceOp":"append"} +{"type":"tool/result","seq":227,"time":1783611707114,"data":{"turn":1,"step":3,"callId":"call_00_GVknJu2tksKkP4lALCwh0926","content":[{"type":"text","text":"The file /private/var/folders/2c/psb0_fmx7hbgz558xjt_f0l00000gn/T/acp-snap-cwd-QzoqnB/settings.txt has been updated successfully."}],"isError":false,"meta":{"diffs":[{"path":"settings.txt","oldText":"color: blue","newText":"color: green"}]}},"sourceEventSeqs":[226],"surfaceOp":"append"} {"type":"step/end","seq":228,"time":1783611707114,"data":{"turn":1,"step":3}} {"type":"step/start","seq":229,"time":1783611707114,"data":{"turn":1,"step":4}} {"type":"assistant/chunk","seq":230,"time":1783611707747,"data":{"turn":1,"step":4,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}} diff --git a/examples/acp-agent/tests/snapshots/fs-write-overwrite/session.jsonl b/examples/acp-agent/tests/snapshots/fs-write-overwrite/session.jsonl index 5a3fc5696b..f3db3493b5 100644 --- a/examples/acp-agent/tests/snapshots/fs-write-overwrite/session.jsonl +++ b/examples/acp-agent/tests/snapshots/fs-write-overwrite/session.jsonl @@ -115,7 +115,7 @@ {"type":"assistant/chunk","seq":113,"time":1783352094988,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":114,"time":1783352094988,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The file contains \"original contents\". Now I'll replace it with \"replaced\"."},{"type":"tool-call","id":"call_00_N23EvXjDo4c8enyWpIUq4043","name":"write","arguments":"{\"file_path\": \"data.txt\", \"content\": \"replaced\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":228,"outputTokens":79,"cacheReadTokens":2816,"reasoningTokens":17}},"sourceEventSeqs":[70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113],"surfaceOp":"append"} {"type":"tool/call","seq":115,"time":1783352094988,"data":{"turn":1,"step":2,"callId":"call_00_N23EvXjDo4c8enyWpIUq4043","name":"write","arguments":"{\"file_path\": \"data.txt\", \"content\": \"replaced\"}"}} -{"type":"tool/result","seq":116,"time":1783352094995,"data":{"turn":1,"step":2,"callId":"call_00_N23EvXjDo4c8enyWpIUq4043","content":[{"type":"text","text":"/tmp/acp-snap-cwd-hH2sGY/data.txt\nfile\n\nUpdated file\n"}],"isError":false,"meta":{"diffs":[{"path":"data.txt","oldText":"original contents","newText":"replaced"}]}},"sourceEventSeqs":[115],"surfaceOp":"append"} +{"type":"tool/result","seq":116,"time":1783352094995,"data":{"turn":1,"step":2,"callId":"call_00_N23EvXjDo4c8enyWpIUq4043","content":[{"type":"text","text":"/private/tmp/acp-snap-cwd-hH2sGY/data.txt\nfile\n\nUpdated file\n"}],"isError":false,"meta":{"diffs":[{"path":"data.txt","oldText":"original contents","newText":"replaced"}]}},"sourceEventSeqs":[115],"surfaceOp":"append"} {"type":"step/end","seq":117,"time":1783352094995,"data":{"turn":1,"step":2}} {"type":"step/start","seq":118,"time":1783352094995,"data":{"turn":1,"step":3}} {"type":"assistant/chunk","seq":119,"time":1783352096090,"data":{"turn":1,"step":3,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}} diff --git a/examples/acp-agent/tests/snapshots/hook-cc-pretool-ask/session.jsonl b/examples/acp-agent/tests/snapshots/hook-cc-pretool-ask/session.jsonl index 888f2f5c13..810083f2a9 100644 --- a/examples/acp-agent/tests/snapshots/hook-cc-pretool-ask/session.jsonl +++ b/examples/acp-agent/tests/snapshots/hook-cc-pretool-ask/session.jsonl @@ -56,8 +56,8 @@ {"type":"tool/call","seq":54,"time":1783352172557,"data":{"turn":1,"step":1,"callId":"call_00_6k0oGSliVHxGSgqBmMEO4311","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Echo HELLO\"}"}} {"type":"hook/invoked","seq":55,"time":1783352172558,"data":{"turn":1,"point":"PreToolUse","dialect":"claude","handlerId":"claude:PreToolUse:1","matcher":"bash"}} {"type":"hook/result","seq":56,"time":1783352172573,"data":{"turn":1,"point":"PreToolUse","handlerId":"claude:PreToolUse:1","decision":"ask","exitCode":0,"durationMs":14.113374999999905}} -{"type":"approval/asked","seq":57,"time":1783962235813,"data":{"id":"97616288-1a5e-4110-a75d-7616a24adcc4","toolName":"bash","callId":"call_00_6k0oGSliVHxGSgqBmMEO4311","reason":"bash requires manual approval in this session"}} -{"type":"approval/decided","seq":58,"time":1783962235813,"data":{"id":"97616288-1a5e-4110-a75d-7616a24adcc4","outcome":"rejected"}} +{"type":"approval/asked","seq":57,"time":1783962235813,"data":{"id":"e629ff8c-60d2-49b1-9603-333b16ef3306","toolName":"bash","callId":"call_00_6k0oGSliVHxGSgqBmMEO4311","reason":"bash requires manual approval in this session"}} +{"type":"approval/decided","seq":58,"time":1783962235813,"data":{"id":"e629ff8c-60d2-49b1-9603-333b16ef3306","outcome":"rejected"}} {"type":"tool/result","seq":59,"time":1783962235814,"data":{"turn":1,"step":1,"callId":"call_00_6k0oGSliVHxGSgqBmMEO4311","content":[{"type":"text","text":"Error: the user rejected tool \"bash\""}],"isError":true},"sourceEventSeqs":[54],"surfaceOp":"append"} {"type":"step/end","seq":60,"time":1783962235814,"data":{"turn":1,"step":1}} {"type":"step/start","seq":61,"time":1783962235814,"data":{"turn":1,"step":2}} diff --git a/examples/acp-agent/tests/snapshots/lsp-definition/tool-schemas.expected.json b/examples/acp-agent/tests/snapshots/lsp-definition/tool-schemas.expected.json index b42a434388..8359d03017 100644 --- a/examples/acp-agent/tests/snapshots/lsp-definition/tool-schemas.expected.json +++ b/examples/acp-agent/tests/snapshots/lsp-definition/tool-schemas.expected.json @@ -434,7 +434,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -459,15 +459,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective; valid only with action edit." + "description": "Replacement objective for action edit; ignored for other actions." }, "max_goal_rounds": { "type": "number", - "description": "Replacement cap; valid only with action edit." + "description": "Replacement round cap for action edit; ignored for other actions." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition; required only with action blocked." + "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." } }, "required": [ diff --git a/examples/acp-agent/tests/snapshots/model-switching/tool-schemas.expected.json b/examples/acp-agent/tests/snapshots/model-switching/tool-schemas.expected.json index ef40784fa5..dac439b700 100644 --- a/examples/acp-agent/tests/snapshots/model-switching/tool-schemas.expected.json +++ b/examples/acp-agent/tests/snapshots/model-switching/tool-schemas.expected.json @@ -397,7 +397,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -422,15 +422,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective; valid only with action edit." + "description": "Replacement objective for action edit; ignored for other actions." }, "max_goal_rounds": { "type": "number", - "description": "Replacement cap; valid only with action edit." + "description": "Replacement round cap for action edit; ignored for other actions." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition; required only with action blocked." + "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." } }, "required": [ @@ -947,7 +947,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -972,15 +972,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective; valid only with action edit." + "description": "Replacement objective for action edit; ignored for other actions." }, "max_goal_rounds": { "type": "number", - "description": "Replacement cap; valid only with action edit." + "description": "Replacement round cap for action edit; ignored for other actions." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition; required only with action blocked." + "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." } }, "required": [ diff --git a/examples/acp-agent/tests/snapshots/permission-switching/tool-schemas.expected.json b/examples/acp-agent/tests/snapshots/permission-switching/tool-schemas.expected.json index ef40784fa5..dac439b700 100644 --- a/examples/acp-agent/tests/snapshots/permission-switching/tool-schemas.expected.json +++ b/examples/acp-agent/tests/snapshots/permission-switching/tool-schemas.expected.json @@ -397,7 +397,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -422,15 +422,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective; valid only with action edit." + "description": "Replacement objective for action edit; ignored for other actions." }, "max_goal_rounds": { "type": "number", - "description": "Replacement cap; valid only with action edit." + "description": "Replacement round cap for action edit; ignored for other actions." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition; required only with action blocked." + "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." } }, "required": [ @@ -947,7 +947,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -972,15 +972,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective; valid only with action edit." + "description": "Replacement objective for action edit; ignored for other actions." }, "max_goal_rounds": { "type": "number", - "description": "Replacement cap; valid only with action edit." + "description": "Replacement round cap for action edit; ignored for other actions." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition; required only with action blocked." + "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." } }, "required": [ diff --git a/examples/acp-agent/tests/snapshots/plan-mode/tool-schemas.expected.json b/examples/acp-agent/tests/snapshots/plan-mode/tool-schemas.expected.json index ef40784fa5..dac439b700 100644 --- a/examples/acp-agent/tests/snapshots/plan-mode/tool-schemas.expected.json +++ b/examples/acp-agent/tests/snapshots/plan-mode/tool-schemas.expected.json @@ -397,7 +397,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -422,15 +422,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective; valid only with action edit." + "description": "Replacement objective for action edit; ignored for other actions." }, "max_goal_rounds": { "type": "number", - "description": "Replacement cap; valid only with action edit." + "description": "Replacement round cap for action edit; ignored for other actions." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition; required only with action blocked." + "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." } }, "required": [ @@ -947,7 +947,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -972,15 +972,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective; valid only with action edit." + "description": "Replacement objective for action edit; ignored for other actions." }, "max_goal_rounds": { "type": "number", - "description": "Replacement cap; valid only with action edit." + "description": "Replacement round cap for action edit; ignored for other actions." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition; required only with action blocked." + "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." } }, "required": [ diff --git a/examples/acp-agent/tests/snapshots/pty-tools/tool-schemas.expected.json b/examples/acp-agent/tests/snapshots/pty-tools/tool-schemas.expected.json index 529b1419da..0ba75afbdc 100644 --- a/examples/acp-agent/tests/snapshots/pty-tools/tool-schemas.expected.json +++ b/examples/acp-agent/tests/snapshots/pty-tools/tool-schemas.expected.json @@ -526,7 +526,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -551,15 +551,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective; valid only with action edit." + "description": "Replacement objective for action edit; ignored for other actions." }, "max_goal_rounds": { "type": "number", - "description": "Replacement cap; valid only with action edit." + "description": "Replacement round cap for action edit; ignored for other actions." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition; required only with action blocked." + "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." } }, "required": [ diff --git a/examples/acp-agent/tests/snapshots/skill-load/tool-schemas.expected.json b/examples/acp-agent/tests/snapshots/skill-load/tool-schemas.expected.json index b01e7683d1..f3739bbc42 100644 --- a/examples/acp-agent/tests/snapshots/skill-load/tool-schemas.expected.json +++ b/examples/acp-agent/tests/snapshots/skill-load/tool-schemas.expected.json @@ -397,7 +397,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -422,15 +422,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective; valid only with action edit." + "description": "Replacement objective for action edit; ignored for other actions." }, "max_goal_rounds": { "type": "number", - "description": "Replacement cap; valid only with action edit." + "description": "Replacement round cap for action edit; ignored for other actions." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition; required only with action blocked." + "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." } }, "required": [ diff --git a/examples/acp-agent/tests/snapshots/text-turn/tool-schemas.expected.json b/examples/acp-agent/tests/snapshots/text-turn/tool-schemas.expected.json index b01e7683d1..f3739bbc42 100644 --- a/examples/acp-agent/tests/snapshots/text-turn/tool-schemas.expected.json +++ b/examples/acp-agent/tests/snapshots/text-turn/tool-schemas.expected.json @@ -397,7 +397,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -422,15 +422,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective; valid only with action edit." + "description": "Replacement objective for action edit; ignored for other actions." }, "max_goal_rounds": { "type": "number", - "description": "Replacement cap; valid only with action edit." + "description": "Replacement round cap for action edit; ignored for other actions." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition; required only with action blocked." + "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." } }, "required": [ diff --git a/examples/acp-agent/tests/snapshots/workspace-context/tool-schemas.expected.json b/examples/acp-agent/tests/snapshots/workspace-context/tool-schemas.expected.json index b01e7683d1..f3739bbc42 100644 --- a/examples/acp-agent/tests/snapshots/workspace-context/tool-schemas.expected.json +++ b/examples/acp-agent/tests/snapshots/workspace-context/tool-schemas.expected.json @@ -397,7 +397,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -422,15 +422,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective; valid only with action edit." + "description": "Replacement objective for action edit; ignored for other actions." }, "max_goal_rounds": { "type": "number", - "description": "Replacement cap; valid only with action edit." + "description": "Replacement round cap for action edit; ignored for other actions." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition; required only with action blocked." + "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." } }, "required": [ diff --git a/packages/goal/tool-goal/README.md b/packages/goal/tool-goal/README.md index 3f286f8ec3..4fc7cb0ea9 100644 --- a/packages/goal/tool-goal/README.md +++ b/packages/goal/tool-goal/README.md @@ -6,7 +6,7 @@ The model-facing control surface for [`ctx.goals`](../goal/README.md): `get_goal - `get_goal()` returns the current goal or `null`, including the compare-and-set id/revision, durable phase, admitted/capped goal rounds, any blocker reason, and current process-local activation. - `create_goal(objective, max_goal_rounds?)` creates one goal from a direct top-level human turn. The model may infer long-running goal intent without an exact command phrase; non-human turns and subagents are rejected at execution. -- `update_goal(goal_id, revision, action, objective?, max_goal_rounds?, blocked_reason?)` supports `edit`, `pause`, `resume`, `complete`, and `blocked`. Replacements belong only to `edit`; `blocked_reason` is required only for `blocked` and is persisted with the stable code `model-reported`. +- `update_goal(goal_id, revision, action, objective?, max_goal_rounds?, blocked_reason?)` supports `edit`, `pause`, `resume`, `complete`, and `blocked`. `edit` consumes replacement fields, `blocked` requires and persists `blocked_reason` with the stable code `model-reported`, and every action ignores fields it does not consume so providers that populate every schema property cannot prevent a valid transition. All calls are exclusive, so a model-ordered batch observes earlier mutations and their new revisions. ACP and other clients receive pure generic cards: read for `get_goal`, other for mutations. diff --git a/packages/goal/tool-goal/src/index.ts b/packages/goal/tool-goal/src/index.ts index 009a00376f..5be69a9978 100644 --- a/packages/goal/tool-goal/src/index.ts +++ b/packages/goal/tool-goal/src/index.ts @@ -248,8 +248,9 @@ export function apply(ctx: Context, config: Config): void { name: 'update_goal', description: 'Update the exact current goal revision. edit, pause, and resume require a direct ' + 'top-level human request. During an automatic continuation of the current goal, complete ' - + 'and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains ' - + 'responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.', + + 'and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured ' + + 'minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and ' + + 'must explain it in blocked_reason.', parameters: { goal_id: { type: 'string', required: true, description: 'Exact id returned by get_goal.' }, revision: { type: 'number', required: true, description: 'Exact positive revision returned by get_goal.' }, @@ -259,11 +260,11 @@ export function apply(ctx: Context, config: Config): void { enum: UPDATE_ACTIONS, description: 'edit | pause | resume | complete | blocked', }, - objective: { type: 'string', description: 'Replacement objective; valid only with action edit.' }, - max_goal_rounds: { type: 'number', description: 'Replacement cap; valid only with action edit.' }, + objective: { type: 'string', description: 'Replacement objective for action edit; ignored for other actions.' }, + max_goal_rounds: { type: 'number', description: 'Replacement round cap for action edit; ignored for other actions.' }, blocked_reason: { type: 'string', - description: 'Concrete blocking condition; required only with action blocked.', + description: 'Concrete blocking condition for action blocked; required for blocked, ignored for other actions.', }, }, output: GOAL_OUTPUT, @@ -276,21 +277,12 @@ export function apply(ctx: Context, config: Config): void { } if (args.action === 'edit') { requireDirectHuman(ctx, execution) - if (args.blocked_reason !== undefined) { - throw new HarnessError('blocked_reason is valid only with action blocked', 'GOAL_TOOL_INVALID_UPDATE') - } const goal = ctx.goals.edit(execution.agent, ref, replacements) observeMutation(terminalTurns, execution, false) return Promise.resolve(goalValue(goal)) } if (args.action === 'pause' || args.action === 'resume') { requireDirectHuman(ctx, execution) - if (args.objective !== undefined || args.max_goal_rounds !== undefined || args.blocked_reason !== undefined) { - throw new HarnessError( - 'objective and max_goal_rounds are valid only with action edit; blocked_reason is valid only with action blocked', - 'GOAL_TOOL_INVALID_UPDATE', - ) - } const goal = args.action === 'pause' ? ctx.goals.pause(execution.agent, ref) : ctx.goals.resume(execution.agent, ref) @@ -298,15 +290,6 @@ export function apply(ctx: Context, config: Config): void { return Promise.resolve(goalValue(goal)) } const authority = completionAuthority(ctx, execution) - if (args.objective !== undefined || args.max_goal_rounds !== undefined) { - throw new HarnessError( - 'objective and max_goal_rounds are valid only with action edit', - 'GOAL_TOOL_INVALID_UPDATE', - ) - } - if (args.action === 'complete' && args.blocked_reason !== undefined) { - throw new HarnessError('blocked_reason is valid only with action blocked', 'GOAL_TOOL_INVALID_UPDATE') - } if (args.action === 'blocked' && (args.blocked_reason === undefined || args.blocked_reason.trim().length === 0)) { throw new HarnessError('blocked_reason is required with action blocked', 'GOAL_TOOL_INVALID_UPDATE') diff --git a/packages/goal/tool-goal/tests/tool-goal.spec.ts b/packages/goal/tool-goal/tests/tool-goal.spec.ts index 7b2ccdc347..30270fd34d 100644 --- a/packages/goal/tool-goal/tests/tool-goal.spec.ts +++ b/packages/goal/tool-goal/tests/tool-goal.spec.ts @@ -377,26 +377,12 @@ describe('goal tool state transitions', () => { closeTurn(root, turn) }) - it('returns structured domain and conditional-argument failures', async () => { + it('returns structured domain and required-argument failures', async () => { const { ctx, root } = await harness() openTurn(root, { kind: 'user' }) const invalidCreate = await execute(ctx, 'create_goal', { objective: ' ' }, root.agent) expect(invalidCreate.error?.info?.code).toBe('GOAL_INVALID_OBJECTIVE') const created = ctx.goals.create(root.agent, { objective: 'valid' }) - const replacement = await execute(ctx, 'update_goal', { - goal_id: created.id, - revision: created.revision, - action: 'pause', - objective: 'not valid for pause', - }, root.agent) - expect(replacement.error?.info?.code).toBe('GOAL_TOOL_INVALID_UPDATE') - const terminalUpdate = await execute(ctx, 'update_goal', { - goal_id: created.id, - revision: created.revision, - action: 'complete', - max_goal_rounds: 2, - }, root.agent) - expect(terminalUpdate.error?.info?.code).toBe('GOAL_TOOL_INVALID_UPDATE') const blockedWithoutReason = await execute(ctx, 'update_goal', { goal_id: created.id, revision: created.revision, action: 'blocked', }, root.agent) @@ -405,24 +391,45 @@ describe('goal tool state transitions', () => { goal_id: created.id, revision: created.revision, action: 'blocked', blocked_reason: ' ', }, root.agent) expect(blockedWithEmptyReason.error?.info?.code).toBe('GOAL_TOOL_INVALID_UPDATE') - const completeWithReason = await execute(ctx, 'update_goal', { - goal_id: created.id, revision: created.revision, action: 'complete', blocked_reason: 'Not a blocker.', - }, root.agent) - expect(completeWithReason.error?.info?.code).toBe('GOAL_TOOL_INVALID_UPDATE') - const editWithReason = await execute(ctx, 'update_goal', { - goal_id: created.id, - revision: created.revision, - action: 'edit', - objective: 'still valid', - blocked_reason: 'Not valid for edit.', - }, root.agent) - expect(editWithReason.error?.info?.code).toBe('GOAL_TOOL_INVALID_UPDATE') const malformedRef = await execute(ctx, 'update_goal', { goal_id: '', revision: 0, action: 'edit', objective: 'x', + max_goal_rounds: 0, blocked_reason: '', }, root.agent) expect(malformedRef.error?.info?.code).toBe('GOAL_TOOL_INVALID_UPDATE') }) + it('ignores unused action fields emitted by strict-schema providers', async () => { + const { ctx, root } = await harness() + openTurn(root, { kind: 'user' }) + let goal = ctx.goals.create(root.agent, { objective: 'valid' }) + const strictArgs = { + objective: '', + max_goal_rounds: 0, + blocked_reason: '', + } + + const paused = await execute(ctx, 'update_goal', { + goal_id: goal.id, revision: goal.revision, action: 'pause', ...strictArgs, + }, root.agent) + goal = ctx.goals.get(root.agent)! + expect(resultGoal(paused)).toMatchObject({ phase: 'paused', objective: 'valid' }) + const resumed = await execute(ctx, 'update_goal', { + goal_id: goal.id, revision: goal.revision, action: 'resume', ...strictArgs, + }, root.agent) + goal = ctx.goals.get(root.agent)! + expect(resultGoal(resumed)).toMatchObject({ phase: 'active', objective: 'valid' }) + const edited = await execute(ctx, 'update_goal', { + goal_id: goal.id, revision: goal.revision, action: 'edit', objective: 'edited', + max_goal_rounds: 7, blocked_reason: '', + }, root.agent) + goal = ctx.goals.get(root.agent)! + expect(resultGoal(edited)).toMatchObject({ objective: 'edited', maxGoalRounds: 7 }) + const complete = await execute(ctx, 'update_goal', { + goal_id: goal.id, revision: goal.revision, action: 'complete', ...strictArgs, + }, root.agent) + expect(resultGoal(complete)).toMatchObject({ phase: 'complete', objective: 'edited', maxGoalRounds: 7 }) + }) + it('allows exact goal rounds to complete but not edit or pause', async () => { const { ctx, root } = await harness() const humanTurn = openTurn(root, { kind: 'user' }) From e27a1319ef2d1c60d0d25c96fe216b74919ae505 Mon Sep 17 00:00:00 2001 From: Turtle Date: Fri, 24 Jul 2026 12:59:06 +0800 Subject: [PATCH 2/8] test(goal): align master snapshots --- docs/tool-catalog.md | 20 +++++++------------ .../system-prompt.expected.md | 8 ++++---- .../tests/snapshots/bash-spill/session.jsonl | 2 +- .../both-mode-turn/system-prompt.expected.md | 8 ++++---- .../code-mode-turn/system-prompt.expected.md | 8 ++++---- .../system-prompt.expected.md | 8 ++++---- .../escalation-approved/session.jsonl | 4 ++-- .../escalation-rejected/session.jsonl | 4 ++-- .../fs-escalation-approved/session.jsonl | 6 +++--- .../tests/snapshots/fs-write/session.jsonl | 2 +- .../hook-cc-pretool-ask/session.jsonl | 4 ++-- .../tests/snapshots/plan-mode/session.jsonl | 4 ++-- .../goal/tool-goal/tests/tool-goal.spec.ts | 12 +++++++++-- 13 files changed, 46 insertions(+), 44 deletions(-) diff --git a/docs/tool-catalog.md b/docs/tool-catalog.md index 4f9de3644f..162232c35b 100644 --- a/docs/tool-catalog.md +++ b/docs/tool-catalog.md @@ -48,7 +48,6 @@ Ask the user a concise question when you need confirmation, a choice, or missing "description": "Questions to ask the user before continuing.", "items": { "type": "object", - "additionalProperties": true, "properties": { "id": { "type": "string", @@ -67,7 +66,6 @@ Ask the user a concise question when you need confirmation, a choice, or missing "description": "Optional choices to show the user. If you recommend one, put it first and append \"(Recommended)\" to that label.", "items": { "type": "object", - "additionalProperties": true, "properties": { "label": { "type": "string", @@ -231,7 +229,7 @@ Source: [`packages/cordis/tool-cordis/src/index.ts`](../packages/cordis/tool-cor ### `cordis_mount` -Mount a NEW cordis plugin into the live runtime that is running THIS agent (self-modification). `code` runs as the body of an async JavaScript function in an isolated sandbox and MUST `return` a plugin. Two forms: FUNCTION form `return (ctx) => { … }` — declares no inject, so it can register tools, listen to events, and provide services, but reaching ANY service (e.g. ctx.bash) throws; use it only when you need no services. OBJECT form `return { name?, inject: ['bash', 'llm', …], apply(ctx) { … } }` — declares dependencies, and cordis activates the plugin only after the services exist; PREFER this form. You may reach ONLY the services you list in inject: an undeclared service throws even if it exists, because an undeclared dependency would not be cleaned up if its provider is unmounted. BEFORE calling a service from your code, read cordis_inspect what:"api" — it lists method signatures AND the type shapes of their arguments/returns (do not guess a field's type; e.g. a bash run's stdout is an object, not a string). Inside `apply`, use the standard cordis API: `ctx.on(event, listener)` to observe events (see cordis_inspect what:"events"), or call `harness.registerTool(ctx, harness.defineTool({ name, description, parameters: { text: { type: 'string', required: true } }, output: { schema: { type: 'string' }, render(_args, value) { return [{ type: 'text', text: value }] } }, async execute(args) { return args.text } }))` to give yourself a new tool — it becomes callable on your NEXT step. Tool parameters: each key IS a property — { type: 'string'|'number'|'integer'|'boolean'|'null'|'object'|'array'|'json', required?: true, description?, enum?, const?, items?, properties? }; every direct DSL object declares additionalProperties: true|false, and oneOf: [schema, schema, ...] replaces type for an exact-one union. A raw JSON-Schema { type: 'object', properties, required?: […] } wrapper is also accepted with open-by-default objects. A tool's `execute` MUST return the lossless JSON value declared by `output.schema`; `output.render(args, value)` separately returns Native/model content blocks. Mounts can COMPOSE: one plugin may `ctx.provide('name', value)` a service and another may declare `inject: ['name']` to consume it — the consumer stays pending until the provider exists and returns to pending when the provider is unmounted. Everything registered inside `apply` is cleaned up automatically on unmount. Sandbox globals: `console` (tagged `[cordis:]`, writes through to the harness terminal), `harness.defineTool`, `harness.registerTool`, `btoa`, `atob`, `TextEncoder`, `TextDecoder`. Node APIs are DISABLED — do filesystem/network/timer work through the cordis services, never Node built-ins: `require`, `setTimeout`/`setInterval`, and `fetch` throw redirect errors; `process` and `Buffer` are undefined. Instead use inject: ['fs'] + ctx.fs for files, inject: ['web'] + ctx.web for HTTP, inject: ['bash'] + ctx.bash for processes, and inject: ['timer'] + ctx.setTimeout/ctx.setInterval for timing (fiber effects, auto-cleaned on unmount) — cordis_inspect what:"api" shows what THIS runtime provides. Write PLAIN JavaScript, not TypeScript (no `as`, no type annotations). Cautions: (1) waterfall events (e.g. tools/pre-execute) hand the listener a trailing `next` callback which MUST be called — returning without `next()` VETOES the call; prefer plain notification events unless you intend to intercept. (2) Never await something that only resolves after the current turn (your code runs INSIDE a tool call of that turn — it would deadlock). (3) Your `ctx` is a restricted façade: you can register tools, observe events, provide/consume services, and use timers, but framework internals (ctx.root, ctx.fiber, ctx.extend, ctx.plugin, …) are withheld. It is not a security boundary though — the services you inject (e.g. ctx.bash) reach the real runtime. +Mount a NEW cordis plugin into the live runtime that is running THIS agent (self-modification). `code` runs as the body of an async JavaScript function in an isolated sandbox and MUST `return` a plugin. Two forms: FUNCTION form `return (ctx) => { … }` — declares no inject, so it can register tools, listen to events, and provide services, but reaching ANY service (e.g. ctx.bash) throws; use it only when you need no services. OBJECT form `return { name?, inject: ['bash', 'llm', …], apply(ctx) { … } }` — declares dependencies, and cordis activates the plugin only after the services exist; PREFER this form. You may reach ONLY the services you list in inject: an undeclared service throws even if it exists, because an undeclared dependency would not be cleaned up if its provider is unmounted. BEFORE calling a service from your code, read cordis_inspect what:"api" — it lists method signatures AND the type shapes of their arguments/returns (do not guess a field's type; e.g. a bash run's stdout is an object, not a string). Inside `apply`, use the standard cordis API: `ctx.on(event, listener)` to observe events (see cordis_inspect what:"events"), or call `harness.registerTool(ctx, harness.defineTool({ name, description, parameters: { text: { type: 'string', required: true } }, async execute(args) { … } }))` to give yourself a new tool — it becomes callable on your NEXT step. Tool parameters: each key IS a property — { type: 'string'|'number'|'boolean'|'object'|'array', required?: true, description?, enum?, items?, properties? }; a JSON-Schema-style { type: 'object', properties, required: […] } wrapper and type 'integer' are also accepted and normalized. A tool's `execute` MUST return an ARRAY of content blocks, e.g. `return [{ type: 'text', text: someString }]` — never a bare string. Mounts can COMPOSE: one plugin may `ctx.provide('name', value)` a service and another may declare `inject: ['name']` to consume it — the consumer stays pending until the provider exists and returns to pending when the provider is unmounted. Everything registered inside `apply` is cleaned up automatically on unmount. Sandbox globals: `console` (tagged `[cordis:]`, writes through to the harness terminal), `harness.defineTool`, `harness.registerTool`, `btoa`, `atob`, `TextEncoder`, `TextDecoder`. Node APIs are DISABLED — do filesystem/network/timer work through the cordis services, never Node built-ins: `require`, `setTimeout`/`setInterval`, and `fetch` throw redirect errors; `process` and `Buffer` are undefined. Instead use inject: ['fs'] + ctx.fs for files, inject: ['web'] + ctx.web for HTTP, inject: ['bash'] + ctx.bash for processes, and inject: ['timer'] + ctx.setTimeout/ctx.setInterval for timing (fiber effects, auto-cleaned on unmount) — cordis_inspect what:"api" shows what THIS runtime provides. Write PLAIN JavaScript, not TypeScript (no `as`, no type annotations). Cautions: (1) waterfall events (e.g. tools/pre-execute) hand the listener a trailing `next` callback which MUST be called — returning without `next()` VETOES the call; prefer plain notification events unless you intend to intercept. (2) Never await something that only resolves after the current turn (your code runs INSIDE a tool call of that turn — it would deadlock). (3) Your `ctx` is a restricted façade: you can register tools, observe events, provide/consume services, and use timers, but framework internals (ctx.root, ctx.fiber, ctx.extend, ctx.plugin, …) are withheld. It is not a security boundary though — the services you inject (e.g. ctx.bash) reach the real runtime. ```json { @@ -630,7 +628,7 @@ Source: [`packages/goal/tool-goal/src/index.ts`](../packages/goal/tool-goal/src/ ### `update_goal` -Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason. +Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason. ```json { @@ -657,15 +655,15 @@ Update the exact current goal revision. edit, pause, and resume require a direct }, "objective": { "type": "string", - "description": "Replacement objective; valid only with action edit." + "description": "Replacement objective for action edit; ignored for other actions." }, "max_goal_rounds": { "type": "number", - "description": "Replacement cap; valid only with action edit." + "description": "Replacement round cap for action edit; ignored for other actions." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition; required only with action blocked." + "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." } }, "required": [ @@ -898,7 +896,6 @@ Record and update a structured task list for the current work. Send the ENTIRE l "description": "The COMPLETE task list, replacing any previous list.", "items": { "type": "object", - "additionalProperties": true, "properties": { "content": { "type": "string", @@ -940,7 +937,7 @@ Run a JavaScript workflow script that orchestrates subagents at scale. Use this The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result. Script-body hooks: -- `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. +- `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. @@ -960,7 +957,6 @@ Constraints: concurrency and total-agent caps apply; no filesystem, network, tim "meta": { "type": "object", "description": "The workflow identity block (plain JSON — never code).", - "additionalProperties": true, "properties": { "name": { "type": "string", @@ -979,7 +975,6 @@ Constraints: concurrency and total-agent caps apply; no filesystem, network, tim "description": "Optional phase declarations matched by phase() calls.", "items": { "type": "object", - "additionalProperties": true, "properties": { "title": { "type": "string", @@ -1011,8 +1006,7 @@ Constraints: concurrency and total-agent caps apply; no filesystem, network, tim }, "args": { "type": "object", - "description": "Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}).", - "additionalProperties": true + "description": "Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]})." } }, "required": [ diff --git a/examples/acp-agent/tests/snapshots/advanced-toolchain/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/advanced-toolchain/system-prompt.expected.md index 73c31413bf..cd179628cd 100644 --- a/examples/acp-agent/tests/snapshots/advanced-toolchain/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/advanced-toolchain/system-prompt.expected.md @@ -190,7 +190,7 @@ interface ToolArgsMap { status: "pending" | "in_progress" | "completed"; } & Record)[]; } & Record; - /** Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason. */ + /** Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason. */ update_goal: { /** Exact id returned by get_goal. */ goal_id: string; @@ -198,11 +198,11 @@ interface ToolArgsMap { revision: number; /** edit | pause | resume | complete | blocked */ action: "edit" | "pause" | "resume" | "complete" | "blocked"; - /** Replacement objective; valid only with action edit. */ + /** Replacement objective for action edit; ignored for other actions. */ objective?: string; - /** Replacement cap; valid only with action edit. */ + /** Replacement round cap for action edit; ignored for other actions. */ max_goal_rounds?: number; - /** Concrete blocking condition; required only with action blocked. */ + /** Concrete blocking condition for action blocked; required for blocked, ignored for other actions. */ blocked_reason?: string; } & Record; /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */ diff --git a/examples/acp-agent/tests/snapshots/bash-spill/session.jsonl b/examples/acp-agent/tests/snapshots/bash-spill/session.jsonl index eb4488858e..7493d1907c 100644 --- a/examples/acp-agent/tests/snapshots/bash-spill/session.jsonl +++ b/examples/acp-agent/tests/snapshots/bash-spill/session.jsonl @@ -11,7 +11,7 @@ {"type":"assistant/chunk","seq":9,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":10,"time":0,"data":{"turn":1,"step":1,"content":[{"type":"tool-call","id":"call_spill","name":"bash","arguments":"{\"command\":\"node -e \\\"process.stdout.write('SPILL_START-' + 'x'.repeat(2000) + '-SPILL_END')\\\"\",\"description\":\"Print large deterministic output\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[5,6,7,8,9],"surfaceOp":"append"} {"type":"tool/call","seq":11,"time":0,"data":{"turn":1,"step":1,"callId":"call_spill","name":"bash","arguments":"{\"command\":\"node -e \\\"process.stdout.write('SPILL_START-' + 'x'.repeat(2000) + '-SPILL_END')\\\"\",\"description\":\"Print large deterministic output\"}"}} -{"type":"tool/result","seq":12,"time":0,"data":{"turn":1,"step":1,"callId":"call_spill","content":[{"type":"text","text":"SPILL_START-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx-SPILL_END\n\n(Omitted 1417 bytes. Full formatted result stored at: /tmp/dsh-acp-snap-ee77dff02/session-fbfcf2f560a0/1bddd2b64176-bash.txt. Use read with offset/limit, or grep this path to search within it.)"}],"isError":false},"sourceEventSeqs":[11],"surfaceOp":"append"} +{"type":"tool/result","seq":12,"time":0,"data":{"turn":1,"step":1,"callId":"call_spill","content":[{"type":"text","text":"SPILL_START-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx-SPILL_END\n\n(Omitted 1417 bytes. Full formatted result stored at: /tmp/dsh-acp-snap-ee77dff02/session-02b64799b331/1bc5704fddf2-bash.txt. Use read with offset/limit, or grep this path to search within it.)"}],"isError":false},"sourceEventSeqs":[11],"surfaceOp":"append"} {"type":"step/end","seq":13,"time":0,"data":{"turn":1,"step":1}} {"type":"step/start","seq":14,"time":0,"data":{"turn":1,"step":2}} {"type":"assistant/chunk","seq":15,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":0,"blockType":"text"}}} diff --git a/examples/acp-agent/tests/snapshots/both-mode-turn/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/both-mode-turn/system-prompt.expected.md index 8744029272..ce858b897e 100644 --- a/examples/acp-agent/tests/snapshots/both-mode-turn/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/both-mode-turn/system-prompt.expected.md @@ -173,7 +173,7 @@ interface ToolArgsMap { status: "pending" | "in_progress" | "completed"; } & Record)[]; } & Record; - /** Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason. */ + /** Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason. */ update_goal: { /** Exact id returned by get_goal. */ goal_id: string; @@ -181,11 +181,11 @@ interface ToolArgsMap { revision: number; /** edit | pause | resume | complete | blocked */ action: "edit" | "pause" | "resume" | "complete" | "blocked"; - /** Replacement objective; valid only with action edit. */ + /** Replacement objective for action edit; ignored for other actions. */ objective?: string; - /** Replacement cap; valid only with action edit. */ + /** Replacement round cap for action edit; ignored for other actions. */ max_goal_rounds?: number; - /** Concrete blocking condition; required only with action blocked. */ + /** Concrete blocking condition for action blocked; required for blocked, ignored for other actions. */ blocked_reason?: string; } & Record; /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */ diff --git a/examples/acp-agent/tests/snapshots/code-mode-turn/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/code-mode-turn/system-prompt.expected.md index 8744029272..ce858b897e 100644 --- a/examples/acp-agent/tests/snapshots/code-mode-turn/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/code-mode-turn/system-prompt.expected.md @@ -173,7 +173,7 @@ interface ToolArgsMap { status: "pending" | "in_progress" | "completed"; } & Record)[]; } & Record; - /** Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason. */ + /** Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason. */ update_goal: { /** Exact id returned by get_goal. */ goal_id: string; @@ -181,11 +181,11 @@ interface ToolArgsMap { revision: number; /** edit | pause | resume | complete | blocked */ action: "edit" | "pause" | "resume" | "complete" | "blocked"; - /** Replacement objective; valid only with action edit. */ + /** Replacement objective for action edit; ignored for other actions. */ objective?: string; - /** Replacement cap; valid only with action edit. */ + /** Replacement round cap for action edit; ignored for other actions. */ max_goal_rounds?: number; - /** Concrete blocking condition; required only with action blocked. */ + /** Concrete blocking condition for action blocked; required for blocked, ignored for other actions. */ blocked_reason?: string; } & Record; /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */ diff --git a/examples/acp-agent/tests/snapshots/code-mode-workspace-context/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/code-mode-workspace-context/system-prompt.expected.md index 8744029272..ce858b897e 100644 --- a/examples/acp-agent/tests/snapshots/code-mode-workspace-context/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/code-mode-workspace-context/system-prompt.expected.md @@ -173,7 +173,7 @@ interface ToolArgsMap { status: "pending" | "in_progress" | "completed"; } & Record)[]; } & Record; - /** Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason. */ + /** Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason. */ update_goal: { /** Exact id returned by get_goal. */ goal_id: string; @@ -181,11 +181,11 @@ interface ToolArgsMap { revision: number; /** edit | pause | resume | complete | blocked */ action: "edit" | "pause" | "resume" | "complete" | "blocked"; - /** Replacement objective; valid only with action edit. */ + /** Replacement objective for action edit; ignored for other actions. */ objective?: string; - /** Replacement cap; valid only with action edit. */ + /** Replacement round cap for action edit; ignored for other actions. */ max_goal_rounds?: number; - /** Concrete blocking condition; required only with action blocked. */ + /** Concrete blocking condition for action blocked; required for blocked, ignored for other actions. */ blocked_reason?: string; } & Record; /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */ diff --git a/examples/acp-agent/tests/snapshots/escalation-approved/session.jsonl b/examples/acp-agent/tests/snapshots/escalation-approved/session.jsonl index 6aa7fda0a8..005b4f4c14 100644 --- a/examples/acp-agent/tests/snapshots/escalation-approved/session.jsonl +++ b/examples/acp-agent/tests/snapshots/escalation-approved/session.jsonl @@ -132,8 +132,8 @@ {"type":"assistant/chunk","seq":130,"time":1783962245385,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":131,"time":1783962245385,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run a command with sandbox_permissions set to danger-full-access, no prior run needed, justified as instructed."},{"type":"tool-call","id":"call_00_d0sAHpJ9mYOJi0z7KNy30441","name":"bash","arguments":"{\"command\": \"printf 'escalated\\\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt\", \"description\": \"Write file outside workspace and verify\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to write a file outside the workspace\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":1501,"outputTokens":174,"cacheReadTokens":0,"reasoningTokens":28}},"sourceEventSeqs":[8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130],"surfaceOp":"append"} {"type":"tool/call","seq":132,"time":1783962245385,"data":{"turn":1,"step":1,"callId":"call_00_d0sAHpJ9mYOJi0z7KNy30441","name":"bash","arguments":"{\"command\": \"printf 'escalated\\\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt\", \"description\": \"Write file outside workspace and verify\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to write a file outside the workspace\"}"}} -{"type":"approval/asked","seq":133,"time":1783962245386,"data":{"id":"21634890-d003-489f-b1b5-d38e187561c5","toolName":"bash","callId":"call_00_d0sAHpJ9mYOJi0z7KNy30441","reason":"escalate sandbox to danger-full-access: the user asked to write a file outside the workspace"}} -{"type":"approval/decided","seq":134,"time":1783962245387,"data":{"id":"21634890-d003-489f-b1b5-d38e187561c5","outcome":"allowed-once"}} +{"type":"approval/asked","seq":133,"time":1783962245386,"data":{"id":"bb75e939-221a-48b0-ad67-3da110ecbd31","toolName":"bash","callId":"call_00_d0sAHpJ9mYOJi0z7KNy30441","reason":"escalate sandbox to danger-full-access: the user asked to write a file outside the workspace"}} +{"type":"approval/decided","seq":134,"time":1783962245387,"data":{"id":"bb75e939-221a-48b0-ad67-3da110ecbd31","outcome":"allowed-once"}} {"type":"tool/result","seq":135,"time":1783962245399,"data":{"turn":1,"step":1,"callId":"call_00_d0sAHpJ9mYOJi0z7KNy30441","content":[{"type":"text","text":"escalated\n"}],"isError":false},"sourceEventSeqs":[132],"surfaceOp":"append"} {"type":"step/end","seq":136,"time":1783962245400,"data":{"turn":1,"step":1}} {"type":"step/start","seq":137,"time":1783962245400,"data":{"turn":1,"step":2}} diff --git a/examples/acp-agent/tests/snapshots/escalation-rejected/session.jsonl b/examples/acp-agent/tests/snapshots/escalation-rejected/session.jsonl index 0d384a55ee..c4f5658452 100644 --- a/examples/acp-agent/tests/snapshots/escalation-rejected/session.jsonl +++ b/examples/acp-agent/tests/snapshots/escalation-rejected/session.jsonl @@ -156,8 +156,8 @@ {"type":"assistant/chunk","seq":154,"time":1783962246274,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":155,"time":1783962246274,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run a specific command with `sandbox_permissions` set to `danger-full-access` and a specific justification. They explicitly said NOT to run it without sandbox_permissions first. Let me do exactly that."},{"type":"tool-call","id":"call_00_WB1vnPomi8yr6MlcFKTj7912","name":"bash","arguments":"{\"command\": \"printf 'escalated\\\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt\", \"description\": \"Write to /tmp and verify, then clean up\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to write a file outside the workspace\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":1509,"outputTokens":198,"cacheReadTokens":0,"reasoningTokens":48}},"sourceEventSeqs":[8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154],"surfaceOp":"append"} {"type":"tool/call","seq":156,"time":1783962246274,"data":{"turn":1,"step":1,"callId":"call_00_WB1vnPomi8yr6MlcFKTj7912","name":"bash","arguments":"{\"command\": \"printf 'escalated\\\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt\", \"description\": \"Write to /tmp and verify, then clean up\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to write a file outside the workspace\"}"}} -{"type":"approval/asked","seq":157,"time":1783962246275,"data":{"id":"4ffb95e3-8f94-41db-bfd3-2f56e6631ef8","toolName":"bash","callId":"call_00_WB1vnPomi8yr6MlcFKTj7912","reason":"escalate sandbox to danger-full-access: the user asked to write a file outside the workspace"}} -{"type":"approval/decided","seq":158,"time":1783962246275,"data":{"id":"4ffb95e3-8f94-41db-bfd3-2f56e6631ef8","outcome":"rejected"}} +{"type":"approval/asked","seq":157,"time":1783962246275,"data":{"id":"9a997191-81a0-4076-914b-255ebe3e6882","toolName":"bash","callId":"call_00_WB1vnPomi8yr6MlcFKTj7912","reason":"escalate sandbox to danger-full-access: the user asked to write a file outside the workspace"}} +{"type":"approval/decided","seq":158,"time":1783962246275,"data":{"id":"9a997191-81a0-4076-914b-255ebe3e6882","outcome":"rejected"}} {"type":"tool/result","seq":159,"time":1783962246275,"data":{"turn":1,"step":1,"callId":"call_00_WB1vnPomi8yr6MlcFKTj7912","content":[{"type":"text","text":"Error: the user rejected escalating this command to \"danger-full-access\""}],"isError":true},"sourceEventSeqs":[156],"surfaceOp":"append"} {"type":"step/end","seq":160,"time":1783962246276,"data":{"turn":1,"step":1}} {"type":"step/start","seq":161,"time":1783962246276,"data":{"turn":1,"step":2}} diff --git a/examples/acp-agent/tests/snapshots/fs-escalation-approved/session.jsonl b/examples/acp-agent/tests/snapshots/fs-escalation-approved/session.jsonl index 918b3ab4d5..67f3d15be0 100644 --- a/examples/acp-agent/tests/snapshots/fs-escalation-approved/session.jsonl +++ b/examples/acp-agent/tests/snapshots/fs-escalation-approved/session.jsonl @@ -90,9 +90,9 @@ {"type":"assistant/chunk","seq":88,"time":1784045703776,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":89,"time":1784045703780,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to create a file using the write tool with sandbox_permissions. Let me do that."},{"type":"tool-call","id":"call_00_Fnymmavpr4klMDy4Fdej3227","name":"write","arguments":"{\"file_path\": \"escalated.md\", \"content\": \"escalated\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to escalate this write\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":3871,"outputTokens":132,"cacheReadTokens":0,"reasoningTokens":23}},"sourceEventSeqs":[8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88],"surfaceOp":"append"} {"type":"tool/call","seq":90,"time":1784045703780,"data":{"turn":1,"step":1,"callId":"call_00_Fnymmavpr4klMDy4Fdej3227","name":"write","arguments":"{\"file_path\": \"escalated.md\", \"content\": \"escalated\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to escalate this write\"}"}} -{"type":"approval/asked","seq":91,"time":1784045703782,"data":{"id":"d7250dff-fc08-4163-9b51-f43f08d6f36f","toolName":"write","callId":"call_00_Fnymmavpr4klMDy4Fdej3227","reason":"escalate sandbox to danger-full-access: the user asked to escalate this write"}} -{"type":"approval/decided","seq":92,"time":1784045703786,"data":{"id":"d7250dff-fc08-4163-9b51-f43f08d6f36f","outcome":"allowed-once"}} -{"type":"tool/result","seq":93,"time":1784045703798,"data":{"turn":1,"step":1,"callId":"call_00_Fnymmavpr4klMDy4Fdej3227","content":[{"type":"text","text":"/var/folders/2g/b32ct0qn1d728l_v6tdkjytr0000gn/T/acp-snap-cwd-vmEGzd/escalated.md\nfile\n\nCreated file\n"}],"isError":false,"meta":{"diffs":[]}},"sourceEventSeqs":[90],"surfaceOp":"append"} +{"type":"approval/asked","seq":91,"time":1784045703782,"data":{"id":"f6a28337-87da-4555-bc47-2b5354300cc4","toolName":"write","callId":"call_00_Fnymmavpr4klMDy4Fdej3227","reason":"escalate sandbox to danger-full-access: the user asked to escalate this write"}} +{"type":"approval/decided","seq":92,"time":1784045703786,"data":{"id":"f6a28337-87da-4555-bc47-2b5354300cc4","outcome":"allowed-once"}} +{"type":"tool/result","seq":93,"time":1784045703798,"data":{"turn":1,"step":1,"callId":"call_00_Fnymmavpr4klMDy4Fdej3227","content":[{"type":"text","text":"/private/var/folders/2g/b32ct0qn1d728l_v6tdkjytr0000gn/T/acp-snap-cwd-vmEGzd/escalated.md\nfile\n\nCreated file\n"}],"isError":false,"meta":{"diffs":[]}},"sourceEventSeqs":[90],"surfaceOp":"append"} {"type":"step/end","seq":94,"time":1784045703798,"data":{"turn":1,"step":1}} {"type":"step/start","seq":95,"time":1784045703799,"data":{"turn":1,"step":2}} {"type":"assistant/chunk","seq":96,"time":1784045704512,"data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}} diff --git a/examples/acp-agent/tests/snapshots/fs-write/session.jsonl b/examples/acp-agent/tests/snapshots/fs-write/session.jsonl index dbaaf8a8b6..46de5fa221 100644 --- a/examples/acp-agent/tests/snapshots/fs-write/session.jsonl +++ b/examples/acp-agent/tests/snapshots/fs-write/session.jsonl @@ -63,7 +63,7 @@ {"type":"assistant/chunk","seq":61,"time":1783352079886,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":62,"time":1783352079888,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to create a file named notes.txt with the content \"hello world\" using the write tool, then reply with \"DONE\"."},{"type":"tool-call","id":"call_00_APMUCJJm9lrTSlVbg6dB0185","name":"write","arguments":"{\"file_path\": \"notes.txt\", \"content\": \"hello world\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2891,"outputTokens":92,"cacheReadTokens":0,"reasoningTokens":30}},"sourceEventSeqs":[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],"surfaceOp":"append"} {"type":"tool/call","seq":63,"time":1783352079888,"data":{"turn":1,"step":1,"callId":"call_00_APMUCJJm9lrTSlVbg6dB0185","name":"write","arguments":"{\"file_path\": \"notes.txt\", \"content\": \"hello world\"}"}} -{"type":"tool/result","seq":64,"time":1783352079897,"data":{"turn":1,"step":1,"callId":"call_00_APMUCJJm9lrTSlVbg6dB0185","content":[{"type":"text","text":"/tmp/acp-snap-cwd-sNvn5N/notes.txt\nfile\n\nCreated file\n"}],"isError":false,"meta":{"diffs":[]}},"sourceEventSeqs":[63],"surfaceOp":"append"} +{"type":"tool/result","seq":64,"time":1783352079897,"data":{"turn":1,"step":1,"callId":"call_00_APMUCJJm9lrTSlVbg6dB0185","content":[{"type":"text","text":"/private/tmp/acp-snap-cwd-sNvn5N/notes.txt\nfile\n\nCreated file\n"}],"isError":false,"meta":{"diffs":[]}},"sourceEventSeqs":[63],"surfaceOp":"append"} {"type":"step/end","seq":65,"time":1783352079898,"data":{"turn":1,"step":1}} {"type":"step/start","seq":66,"time":1783352079899,"data":{"turn":1,"step":2}} {"type":"assistant/chunk","seq":67,"time":1783352080825,"data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}} diff --git a/examples/acp-agent/tests/snapshots/hook-cc-pretool-ask/session.jsonl b/examples/acp-agent/tests/snapshots/hook-cc-pretool-ask/session.jsonl index 810083f2a9..baa85c9d77 100644 --- a/examples/acp-agent/tests/snapshots/hook-cc-pretool-ask/session.jsonl +++ b/examples/acp-agent/tests/snapshots/hook-cc-pretool-ask/session.jsonl @@ -56,8 +56,8 @@ {"type":"tool/call","seq":54,"time":1783352172557,"data":{"turn":1,"step":1,"callId":"call_00_6k0oGSliVHxGSgqBmMEO4311","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Echo HELLO\"}"}} {"type":"hook/invoked","seq":55,"time":1783352172558,"data":{"turn":1,"point":"PreToolUse","dialect":"claude","handlerId":"claude:PreToolUse:1","matcher":"bash"}} {"type":"hook/result","seq":56,"time":1783352172573,"data":{"turn":1,"point":"PreToolUse","handlerId":"claude:PreToolUse:1","decision":"ask","exitCode":0,"durationMs":14.113374999999905}} -{"type":"approval/asked","seq":57,"time":1783962235813,"data":{"id":"e629ff8c-60d2-49b1-9603-333b16ef3306","toolName":"bash","callId":"call_00_6k0oGSliVHxGSgqBmMEO4311","reason":"bash requires manual approval in this session"}} -{"type":"approval/decided","seq":58,"time":1783962235813,"data":{"id":"e629ff8c-60d2-49b1-9603-333b16ef3306","outcome":"rejected"}} +{"type":"approval/asked","seq":57,"time":1783962235813,"data":{"id":"c7877002-3d2e-432e-9624-0b9ef3624df3","toolName":"bash","callId":"call_00_6k0oGSliVHxGSgqBmMEO4311","reason":"bash requires manual approval in this session"}} +{"type":"approval/decided","seq":58,"time":1783962235813,"data":{"id":"c7877002-3d2e-432e-9624-0b9ef3624df3","outcome":"rejected"}} {"type":"tool/result","seq":59,"time":1783962235814,"data":{"turn":1,"step":1,"callId":"call_00_6k0oGSliVHxGSgqBmMEO4311","content":[{"type":"text","text":"Error: the user rejected tool \"bash\""}],"isError":true},"sourceEventSeqs":[54],"surfaceOp":"append"} {"type":"step/end","seq":60,"time":1783962235814,"data":{"turn":1,"step":1}} {"type":"step/start","seq":61,"time":1783962235814,"data":{"turn":1,"step":2}} diff --git a/examples/acp-agent/tests/snapshots/plan-mode/session.jsonl b/examples/acp-agent/tests/snapshots/plan-mode/session.jsonl index 71a1189702..7c1b1a7574 100644 --- a/examples/acp-agent/tests/snapshots/plan-mode/session.jsonl +++ b/examples/acp-agent/tests/snapshots/plan-mode/session.jsonl @@ -324,7 +324,7 @@ {"type":"assistant/chunk","seq":322,"time":1784525379646,"data":{"turn":1,"step":3,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":323,"time":1784525379647,"data":{"turn":1,"step":3,"content":[{"type":"reasoning","text":"The plan is approved. Now I need to apply exactly that one edit: change line 2 of notes.txt to say \"hello world\". I'll use the edit tool to replace the current line 2 content."},{"type":"tool-call","id":"call_00_eYuZ6fAuzqr8vtFd3oRg2452","name":"edit","arguments":"{\"file_path\": \"notes.txt\", \"old_string\": \"- the greeting message still says \\\"helo wrld\\\"\", \"new_string\": \"hello world\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":3269,"outputTokens":134,"cacheReadTokens":0,"reasoningTokens":43}},"sourceEventSeqs":[232,233,234,235,236,237,238,239,240,241,242,243,244,245,246,247,248,249,250,251,252,253,254,255,256,257,258,259,260,261,262,263,264,265,266,267,268,269,270,271,272,273,274,275,276,277,278,279,280,281,282,283,284,285,286,287,288,289,290,291,292,293,294,295,296,297,298,299,300,301,302,303,304,305,306,307,308,309,310,311,312,313,314,315,316,317,318,319,320,321,322],"surfaceOp":"append"} {"type":"tool/call","seq":324,"time":1784525379647,"data":{"turn":1,"step":3,"callId":"call_00_eYuZ6fAuzqr8vtFd3oRg2452","name":"edit","arguments":"{\"file_path\": \"notes.txt\", \"old_string\": \"- the greeting message still says \\\"helo wrld\\\"\", \"new_string\": \"hello world\"}"}} -{"type":"tool/result","seq":325,"time":1784525379652,"data":{"turn":1,"step":3,"callId":"call_00_eYuZ6fAuzqr8vtFd3oRg2452","content":[{"type":"text","text":"Error: edit requires reading \"/var/folders/2g/b32ct0qn1d728l_v6tdkjytr0000gn/T/acp-snap-cwd-tia05W/notes.txt\" first"}],"isError":true,"error":{"name":"FsError","code":"FS_NOT_OBSERVED"}},"sourceEventSeqs":[324],"surfaceOp":"append"} +{"type":"tool/result","seq":325,"time":1784525379652,"data":{"turn":1,"step":3,"callId":"call_00_eYuZ6fAuzqr8vtFd3oRg2452","content":[{"type":"text","text":"Error: edit requires reading \"/private/var/folders/2g/b32ct0qn1d728l_v6tdkjytr0000gn/T/acp-snap-cwd-tia05W/notes.txt\" first"}],"isError":true,"error":{"name":"FsError","code":"FS_NOT_OBSERVED"}},"sourceEventSeqs":[324],"surfaceOp":"append"} {"type":"step/end","seq":326,"time":1784525379652,"data":{"turn":1,"step":3}} {"type":"step/start","seq":327,"time":1784525379655,"data":{"turn":1,"step":4}} {"type":"assistant/chunk","seq":328,"time":1784525380205,"data":{"turn":1,"step":4,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}} @@ -428,7 +428,7 @@ {"type":"assistant/chunk","seq":426,"time":1784525381745,"data":{"turn":1,"step":5,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":427,"time":1784525381746,"data":{"turn":1,"step":5,"content":[{"type":"reasoning","text":"Now I can apply the edit."},{"type":"tool-call","id":"call_00_7H6VFdLjBqzzW9M5Z4OR8016","name":"edit","arguments":"{\"file_path\": \"notes.txt\", \"old_string\": \"- the greeting message still says \\\"helo wrld\\\"\", \"new_string\": \"hello world\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":169,"outputTokens":98,"cacheReadTokens":3456,"reasoningTokens":7}},"sourceEventSeqs":[372,373,374,375,376,377,378,379,380,381,382,383,384,385,386,387,388,389,390,391,392,393,394,395,396,397,398,399,400,401,402,403,404,405,406,407,408,409,410,411,412,413,414,415,416,417,418,419,420,421,422,423,424,425,426],"surfaceOp":"append"} {"type":"tool/call","seq":428,"time":1784525381746,"data":{"turn":1,"step":5,"callId":"call_00_7H6VFdLjBqzzW9M5Z4OR8016","name":"edit","arguments":"{\"file_path\": \"notes.txt\", \"old_string\": \"- the greeting message still says \\\"helo wrld\\\"\", \"new_string\": \"hello world\"}"}} -{"type":"tool/result","seq":429,"time":1784525381766,"data":{"turn":1,"step":5,"callId":"call_00_7H6VFdLjBqzzW9M5Z4OR8016","content":[{"type":"text","text":"The file /var/folders/2g/b32ct0qn1d728l_v6tdkjytr0000gn/T/acp-snap-cwd-tia05W/notes.txt has been updated successfully."}],"isError":false,"meta":{"diffs":[{"path":"notes.txt","oldText":"project notes\n- the greeting message still says \"helo wrld\"","newText":"project notes\nhello world"}]}},"sourceEventSeqs":[428],"surfaceOp":"append"} +{"type":"tool/result","seq":429,"time":1784525381766,"data":{"turn":1,"step":5,"callId":"call_00_7H6VFdLjBqzzW9M5Z4OR8016","content":[{"type":"text","text":"The file /private/var/folders/2g/b32ct0qn1d728l_v6tdkjytr0000gn/T/acp-snap-cwd-tia05W/notes.txt has been updated successfully."}],"isError":false,"meta":{"diffs":[{"path":"notes.txt","oldText":"project notes\n- the greeting message still says \"helo wrld\"","newText":"project notes\nhello world"}]}},"sourceEventSeqs":[428],"surfaceOp":"append"} {"type":"step/end","seq":430,"time":1784525381766,"data":{"turn":1,"step":5}} {"type":"step/start","seq":431,"time":1784525381768,"data":{"turn":1,"step":6}} {"type":"assistant/chunk","seq":432,"time":1784525382334,"data":{"turn":1,"step":6,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}} diff --git a/packages/goal/tool-goal/tests/tool-goal.spec.ts b/packages/goal/tool-goal/tests/tool-goal.spec.ts index 30270fd34d..c743f84903 100644 --- a/packages/goal/tool-goal/tests/tool-goal.spec.ts +++ b/packages/goal/tool-goal/tests/tool-goal.spec.ts @@ -409,10 +409,18 @@ describe('goal tool state transitions', () => { } const paused = await execute(ctx, 'update_goal', { - goal_id: goal.id, revision: goal.revision, action: 'pause', ...strictArgs, + goal_id: goal.id, + revision: goal.revision, + action: 'pause', + objective: 'ignored replacement', + max_goal_rounds: 1, + blocked_reason: 'ignored blocker', }, root.agent) goal = ctx.goals.get(root.agent)! - expect(resultGoal(paused)).toMatchObject({ phase: 'paused', objective: 'valid' }) + expect(resultGoal(paused)).toMatchObject({ + phase: 'paused', objective: 'valid', maxGoalRounds: goal.maxGoalRounds, + }) + expect(goal.blockedReason).toBeUndefined() const resumed = await execute(ctx, 'update_goal', { goal_id: goal.id, revision: goal.revision, action: 'resume', ...strictArgs, }, root.agent) From 2e1be7d4d76ac7106c8523ebc1b7844aa4363cc6 Mon Sep 17 00:00:00 2001 From: Turtle Date: Fri, 24 Jul 2026 13:09:13 +0800 Subject: [PATCH 3/8] Revert "test(goal): align master snapshots" This reverts commit 028e63a9eee281a254f113f4721a67343e5521ca. --- docs/tool-catalog.md | 20 ++++++++++++------- .../system-prompt.expected.md | 8 ++++---- .../tests/snapshots/bash-spill/session.jsonl | 2 +- .../both-mode-turn/system-prompt.expected.md | 8 ++++---- .../code-mode-turn/system-prompt.expected.md | 8 ++++---- .../system-prompt.expected.md | 8 ++++---- .../escalation-approved/session.jsonl | 4 ++-- .../escalation-rejected/session.jsonl | 4 ++-- .../fs-escalation-approved/session.jsonl | 6 +++--- .../tests/snapshots/fs-write/session.jsonl | 2 +- .../hook-cc-pretool-ask/session.jsonl | 4 ++-- .../tests/snapshots/plan-mode/session.jsonl | 4 ++-- .../goal/tool-goal/tests/tool-goal.spec.ts | 12 ++--------- 13 files changed, 44 insertions(+), 46 deletions(-) diff --git a/docs/tool-catalog.md b/docs/tool-catalog.md index 162232c35b..4f9de3644f 100644 --- a/docs/tool-catalog.md +++ b/docs/tool-catalog.md @@ -48,6 +48,7 @@ Ask the user a concise question when you need confirmation, a choice, or missing "description": "Questions to ask the user before continuing.", "items": { "type": "object", + "additionalProperties": true, "properties": { "id": { "type": "string", @@ -66,6 +67,7 @@ Ask the user a concise question when you need confirmation, a choice, or missing "description": "Optional choices to show the user. If you recommend one, put it first and append \"(Recommended)\" to that label.", "items": { "type": "object", + "additionalProperties": true, "properties": { "label": { "type": "string", @@ -229,7 +231,7 @@ Source: [`packages/cordis/tool-cordis/src/index.ts`](../packages/cordis/tool-cor ### `cordis_mount` -Mount a NEW cordis plugin into the live runtime that is running THIS agent (self-modification). `code` runs as the body of an async JavaScript function in an isolated sandbox and MUST `return` a plugin. Two forms: FUNCTION form `return (ctx) => { … }` — declares no inject, so it can register tools, listen to events, and provide services, but reaching ANY service (e.g. ctx.bash) throws; use it only when you need no services. OBJECT form `return { name?, inject: ['bash', 'llm', …], apply(ctx) { … } }` — declares dependencies, and cordis activates the plugin only after the services exist; PREFER this form. You may reach ONLY the services you list in inject: an undeclared service throws even if it exists, because an undeclared dependency would not be cleaned up if its provider is unmounted. BEFORE calling a service from your code, read cordis_inspect what:"api" — it lists method signatures AND the type shapes of their arguments/returns (do not guess a field's type; e.g. a bash run's stdout is an object, not a string). Inside `apply`, use the standard cordis API: `ctx.on(event, listener)` to observe events (see cordis_inspect what:"events"), or call `harness.registerTool(ctx, harness.defineTool({ name, description, parameters: { text: { type: 'string', required: true } }, async execute(args) { … } }))` to give yourself a new tool — it becomes callable on your NEXT step. Tool parameters: each key IS a property — { type: 'string'|'number'|'boolean'|'object'|'array', required?: true, description?, enum?, items?, properties? }; a JSON-Schema-style { type: 'object', properties, required: […] } wrapper and type 'integer' are also accepted and normalized. A tool's `execute` MUST return an ARRAY of content blocks, e.g. `return [{ type: 'text', text: someString }]` — never a bare string. Mounts can COMPOSE: one plugin may `ctx.provide('name', value)` a service and another may declare `inject: ['name']` to consume it — the consumer stays pending until the provider exists and returns to pending when the provider is unmounted. Everything registered inside `apply` is cleaned up automatically on unmount. Sandbox globals: `console` (tagged `[cordis:]`, writes through to the harness terminal), `harness.defineTool`, `harness.registerTool`, `btoa`, `atob`, `TextEncoder`, `TextDecoder`. Node APIs are DISABLED — do filesystem/network/timer work through the cordis services, never Node built-ins: `require`, `setTimeout`/`setInterval`, and `fetch` throw redirect errors; `process` and `Buffer` are undefined. Instead use inject: ['fs'] + ctx.fs for files, inject: ['web'] + ctx.web for HTTP, inject: ['bash'] + ctx.bash for processes, and inject: ['timer'] + ctx.setTimeout/ctx.setInterval for timing (fiber effects, auto-cleaned on unmount) — cordis_inspect what:"api" shows what THIS runtime provides. Write PLAIN JavaScript, not TypeScript (no `as`, no type annotations). Cautions: (1) waterfall events (e.g. tools/pre-execute) hand the listener a trailing `next` callback which MUST be called — returning without `next()` VETOES the call; prefer plain notification events unless you intend to intercept. (2) Never await something that only resolves after the current turn (your code runs INSIDE a tool call of that turn — it would deadlock). (3) Your `ctx` is a restricted façade: you can register tools, observe events, provide/consume services, and use timers, but framework internals (ctx.root, ctx.fiber, ctx.extend, ctx.plugin, …) are withheld. It is not a security boundary though — the services you inject (e.g. ctx.bash) reach the real runtime. +Mount a NEW cordis plugin into the live runtime that is running THIS agent (self-modification). `code` runs as the body of an async JavaScript function in an isolated sandbox and MUST `return` a plugin. Two forms: FUNCTION form `return (ctx) => { … }` — declares no inject, so it can register tools, listen to events, and provide services, but reaching ANY service (e.g. ctx.bash) throws; use it only when you need no services. OBJECT form `return { name?, inject: ['bash', 'llm', …], apply(ctx) { … } }` — declares dependencies, and cordis activates the plugin only after the services exist; PREFER this form. You may reach ONLY the services you list in inject: an undeclared service throws even if it exists, because an undeclared dependency would not be cleaned up if its provider is unmounted. BEFORE calling a service from your code, read cordis_inspect what:"api" — it lists method signatures AND the type shapes of their arguments/returns (do not guess a field's type; e.g. a bash run's stdout is an object, not a string). Inside `apply`, use the standard cordis API: `ctx.on(event, listener)` to observe events (see cordis_inspect what:"events"), or call `harness.registerTool(ctx, harness.defineTool({ name, description, parameters: { text: { type: 'string', required: true } }, output: { schema: { type: 'string' }, render(_args, value) { return [{ type: 'text', text: value }] } }, async execute(args) { return args.text } }))` to give yourself a new tool — it becomes callable on your NEXT step. Tool parameters: each key IS a property — { type: 'string'|'number'|'integer'|'boolean'|'null'|'object'|'array'|'json', required?: true, description?, enum?, const?, items?, properties? }; every direct DSL object declares additionalProperties: true|false, and oneOf: [schema, schema, ...] replaces type for an exact-one union. A raw JSON-Schema { type: 'object', properties, required?: […] } wrapper is also accepted with open-by-default objects. A tool's `execute` MUST return the lossless JSON value declared by `output.schema`; `output.render(args, value)` separately returns Native/model content blocks. Mounts can COMPOSE: one plugin may `ctx.provide('name', value)` a service and another may declare `inject: ['name']` to consume it — the consumer stays pending until the provider exists and returns to pending when the provider is unmounted. Everything registered inside `apply` is cleaned up automatically on unmount. Sandbox globals: `console` (tagged `[cordis:]`, writes through to the harness terminal), `harness.defineTool`, `harness.registerTool`, `btoa`, `atob`, `TextEncoder`, `TextDecoder`. Node APIs are DISABLED — do filesystem/network/timer work through the cordis services, never Node built-ins: `require`, `setTimeout`/`setInterval`, and `fetch` throw redirect errors; `process` and `Buffer` are undefined. Instead use inject: ['fs'] + ctx.fs for files, inject: ['web'] + ctx.web for HTTP, inject: ['bash'] + ctx.bash for processes, and inject: ['timer'] + ctx.setTimeout/ctx.setInterval for timing (fiber effects, auto-cleaned on unmount) — cordis_inspect what:"api" shows what THIS runtime provides. Write PLAIN JavaScript, not TypeScript (no `as`, no type annotations). Cautions: (1) waterfall events (e.g. tools/pre-execute) hand the listener a trailing `next` callback which MUST be called — returning without `next()` VETOES the call; prefer plain notification events unless you intend to intercept. (2) Never await something that only resolves after the current turn (your code runs INSIDE a tool call of that turn — it would deadlock). (3) Your `ctx` is a restricted façade: you can register tools, observe events, provide/consume services, and use timers, but framework internals (ctx.root, ctx.fiber, ctx.extend, ctx.plugin, …) are withheld. It is not a security boundary though — the services you inject (e.g. ctx.bash) reach the real runtime. ```json { @@ -628,7 +630,7 @@ Source: [`packages/goal/tool-goal/src/index.ts`](../packages/goal/tool-goal/src/ ### `update_goal` -Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason. +Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason. ```json { @@ -655,15 +657,15 @@ Update the exact current goal revision. edit, pause, and resume require a direct }, "objective": { "type": "string", - "description": "Replacement objective for action edit; ignored for other actions." + "description": "Replacement objective; valid only with action edit." }, "max_goal_rounds": { "type": "number", - "description": "Replacement round cap for action edit; ignored for other actions." + "description": "Replacement cap; valid only with action edit." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." + "description": "Concrete blocking condition; required only with action blocked." } }, "required": [ @@ -896,6 +898,7 @@ Record and update a structured task list for the current work. Send the ENTIRE l "description": "The COMPLETE task list, replacing any previous list.", "items": { "type": "object", + "additionalProperties": true, "properties": { "content": { "type": "string", @@ -937,7 +940,7 @@ Run a JavaScript workflow script that orchestrates subagents at scale. Use this The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result. Script-body hooks: -- `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const — no oneOf/pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. +- `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. @@ -957,6 +960,7 @@ Constraints: concurrency and total-agent caps apply; no filesystem, network, tim "meta": { "type": "object", "description": "The workflow identity block (plain JSON — never code).", + "additionalProperties": true, "properties": { "name": { "type": "string", @@ -975,6 +979,7 @@ Constraints: concurrency and total-agent caps apply; no filesystem, network, tim "description": "Optional phase declarations matched by phase() calls.", "items": { "type": "object", + "additionalProperties": true, "properties": { "title": { "type": "string", @@ -1006,7 +1011,8 @@ Constraints: concurrency and total-agent caps apply; no filesystem, network, tim }, "args": { "type": "object", - "description": "Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]})." + "description": "Optional JSON input exposed to the script as the `args` global (wrap a bare list as a field, e.g. {\"files\": [...]}).", + "additionalProperties": true } }, "required": [ diff --git a/examples/acp-agent/tests/snapshots/advanced-toolchain/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/advanced-toolchain/system-prompt.expected.md index cd179628cd..73c31413bf 100644 --- a/examples/acp-agent/tests/snapshots/advanced-toolchain/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/advanced-toolchain/system-prompt.expected.md @@ -190,7 +190,7 @@ interface ToolArgsMap { status: "pending" | "in_progress" | "completed"; } & Record)[]; } & Record; - /** Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason. */ + /** Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason. */ update_goal: { /** Exact id returned by get_goal. */ goal_id: string; @@ -198,11 +198,11 @@ interface ToolArgsMap { revision: number; /** edit | pause | resume | complete | blocked */ action: "edit" | "pause" | "resume" | "complete" | "blocked"; - /** Replacement objective for action edit; ignored for other actions. */ + /** Replacement objective; valid only with action edit. */ objective?: string; - /** Replacement round cap for action edit; ignored for other actions. */ + /** Replacement cap; valid only with action edit. */ max_goal_rounds?: number; - /** Concrete blocking condition for action blocked; required for blocked, ignored for other actions. */ + /** Concrete blocking condition; required only with action blocked. */ blocked_reason?: string; } & Record; /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */ diff --git a/examples/acp-agent/tests/snapshots/bash-spill/session.jsonl b/examples/acp-agent/tests/snapshots/bash-spill/session.jsonl index 7493d1907c..eb4488858e 100644 --- a/examples/acp-agent/tests/snapshots/bash-spill/session.jsonl +++ b/examples/acp-agent/tests/snapshots/bash-spill/session.jsonl @@ -11,7 +11,7 @@ {"type":"assistant/chunk","seq":9,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":10,"time":0,"data":{"turn":1,"step":1,"content":[{"type":"tool-call","id":"call_spill","name":"bash","arguments":"{\"command\":\"node -e \\\"process.stdout.write('SPILL_START-' + 'x'.repeat(2000) + '-SPILL_END')\\\"\",\"description\":\"Print large deterministic output\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":10,"outputTokens":5}},"sourceEventSeqs":[5,6,7,8,9],"surfaceOp":"append"} {"type":"tool/call","seq":11,"time":0,"data":{"turn":1,"step":1,"callId":"call_spill","name":"bash","arguments":"{\"command\":\"node -e \\\"process.stdout.write('SPILL_START-' + 'x'.repeat(2000) + '-SPILL_END')\\\"\",\"description\":\"Print large deterministic output\"}"}} -{"type":"tool/result","seq":12,"time":0,"data":{"turn":1,"step":1,"callId":"call_spill","content":[{"type":"text","text":"SPILL_START-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx-SPILL_END\n\n(Omitted 1417 bytes. Full formatted result stored at: /tmp/dsh-acp-snap-ee77dff02/session-02b64799b331/1bc5704fddf2-bash.txt. Use read with offset/limit, or grep this path to search within it.)"}],"isError":false},"sourceEventSeqs":[11],"surfaceOp":"append"} +{"type":"tool/result","seq":12,"time":0,"data":{"turn":1,"step":1,"callId":"call_spill","content":[{"type":"text","text":"SPILL_START-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx-SPILL_END\n\n(Omitted 1417 bytes. Full formatted result stored at: /tmp/dsh-acp-snap-ee77dff02/session-fbfcf2f560a0/1bddd2b64176-bash.txt. Use read with offset/limit, or grep this path to search within it.)"}],"isError":false},"sourceEventSeqs":[11],"surfaceOp":"append"} {"type":"step/end","seq":13,"time":0,"data":{"turn":1,"step":1}} {"type":"step/start","seq":14,"time":0,"data":{"turn":1,"step":2}} {"type":"assistant/chunk","seq":15,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":0,"blockType":"text"}}} diff --git a/examples/acp-agent/tests/snapshots/both-mode-turn/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/both-mode-turn/system-prompt.expected.md index ce858b897e..8744029272 100644 --- a/examples/acp-agent/tests/snapshots/both-mode-turn/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/both-mode-turn/system-prompt.expected.md @@ -173,7 +173,7 @@ interface ToolArgsMap { status: "pending" | "in_progress" | "completed"; } & Record)[]; } & Record; - /** Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason. */ + /** Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason. */ update_goal: { /** Exact id returned by get_goal. */ goal_id: string; @@ -181,11 +181,11 @@ interface ToolArgsMap { revision: number; /** edit | pause | resume | complete | blocked */ action: "edit" | "pause" | "resume" | "complete" | "blocked"; - /** Replacement objective for action edit; ignored for other actions. */ + /** Replacement objective; valid only with action edit. */ objective?: string; - /** Replacement round cap for action edit; ignored for other actions. */ + /** Replacement cap; valid only with action edit. */ max_goal_rounds?: number; - /** Concrete blocking condition for action blocked; required for blocked, ignored for other actions. */ + /** Concrete blocking condition; required only with action blocked. */ blocked_reason?: string; } & Record; /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */ diff --git a/examples/acp-agent/tests/snapshots/code-mode-turn/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/code-mode-turn/system-prompt.expected.md index ce858b897e..8744029272 100644 --- a/examples/acp-agent/tests/snapshots/code-mode-turn/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/code-mode-turn/system-prompt.expected.md @@ -173,7 +173,7 @@ interface ToolArgsMap { status: "pending" | "in_progress" | "completed"; } & Record)[]; } & Record; - /** Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason. */ + /** Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason. */ update_goal: { /** Exact id returned by get_goal. */ goal_id: string; @@ -181,11 +181,11 @@ interface ToolArgsMap { revision: number; /** edit | pause | resume | complete | blocked */ action: "edit" | "pause" | "resume" | "complete" | "blocked"; - /** Replacement objective for action edit; ignored for other actions. */ + /** Replacement objective; valid only with action edit. */ objective?: string; - /** Replacement round cap for action edit; ignored for other actions. */ + /** Replacement cap; valid only with action edit. */ max_goal_rounds?: number; - /** Concrete blocking condition for action blocked; required for blocked, ignored for other actions. */ + /** Concrete blocking condition; required only with action blocked. */ blocked_reason?: string; } & Record; /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */ diff --git a/examples/acp-agent/tests/snapshots/code-mode-workspace-context/system-prompt.expected.md b/examples/acp-agent/tests/snapshots/code-mode-workspace-context/system-prompt.expected.md index ce858b897e..8744029272 100644 --- a/examples/acp-agent/tests/snapshots/code-mode-workspace-context/system-prompt.expected.md +++ b/examples/acp-agent/tests/snapshots/code-mode-workspace-context/system-prompt.expected.md @@ -173,7 +173,7 @@ interface ToolArgsMap { status: "pending" | "in_progress" | "completed"; } & Record)[]; } & Record; - /** Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason. */ + /** Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason. */ update_goal: { /** Exact id returned by get_goal. */ goal_id: string; @@ -181,11 +181,11 @@ interface ToolArgsMap { revision: number; /** edit | pause | resume | complete | blocked */ action: "edit" | "pause" | "resume" | "complete" | "blocked"; - /** Replacement objective for action edit; ignored for other actions. */ + /** Replacement objective; valid only with action edit. */ objective?: string; - /** Replacement round cap for action edit; ignored for other actions. */ + /** Replacement cap; valid only with action edit. */ max_goal_rounds?: number; - /** Concrete blocking condition for action blocked; required for blocked, ignored for other actions. */ + /** Concrete blocking condition; required only with action blocked. */ blocked_reason?: string; } & Record; /** Run a JavaScript workflow script that orchestrates subagents at scale. Use this for work that fans out across many independent pieces — an audit over many files, a migration, multi-angle research, adversarial verification of findings — where you write the orchestration as a script instead of delegating turn by turn. The workflow's identity rides the `meta` parameter as JSON: required `name` (short kebab-case) and `description` strings, optional `whenToUse` string and `phases` array (`{title, detail?, provider?, model?}`). The `script` parameter is the plain JavaScript body ONLY (NOT TypeScript, and NO `export const meta` statement — meta is a parameter, not code), running with top-level await; end with `return ` — the value must be JSON-serializable and is this tool's result. Script-body hooks: - `agent(prompt, opts?): Promise` — run one subagent to completion. Without `opts.schema` it resolves to the child's final text; with `opts.schema` (an object-rooted JSON Schema using ONLY type/properties/required/additionalProperties/items/enum/const/oneOf — no pattern/format/numeric bounds) it resolves to the validated object. Resolves `null` when the child fails (filter with `.filter(Boolean)`). Other opts: `label` (display), `phase` (progress group), and independent `provider`/`model` LLM target overrides (either may be provided alone). Anything else (`effort`/`isolation`/`agentType`) is rejected loudly. - `pipeline(items, ...stages): Promise` — run each item through the stages independently with NO barrier between stages (prefer this for multi-stage work). Each stage receives `(prev, item, index)`. An ordinary stage throw drops that ITEM to `null` and skips its remaining stages. - `parallel(thunks): Promise` — run zero-argument functions concurrently and await ALL of them (a barrier; use only when a stage genuinely needs every prior result together). A throwing thunk resolves to `null`. - `phase(title)` — start a progress phase; `log(message)` — narrate progress; `args` — the tool call's `args` input, verbatim. Misused hooks (bad arguments, unknown options, unsupported schemas, tripped caps) throw errors that ALWAYS kill the script — they never dissolve into a per-item `null`. Constraints: concurrency and total-agent caps apply; no filesystem, network, timers, or Node.js APIs are provided — the agents do the work, the script only coordinates them. The run executes in the foreground: this call returns when the whole script finishes. */ diff --git a/examples/acp-agent/tests/snapshots/escalation-approved/session.jsonl b/examples/acp-agent/tests/snapshots/escalation-approved/session.jsonl index 005b4f4c14..6aa7fda0a8 100644 --- a/examples/acp-agent/tests/snapshots/escalation-approved/session.jsonl +++ b/examples/acp-agent/tests/snapshots/escalation-approved/session.jsonl @@ -132,8 +132,8 @@ {"type":"assistant/chunk","seq":130,"time":1783962245385,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":131,"time":1783962245385,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run a command with sandbox_permissions set to danger-full-access, no prior run needed, justified as instructed."},{"type":"tool-call","id":"call_00_d0sAHpJ9mYOJi0z7KNy30441","name":"bash","arguments":"{\"command\": \"printf 'escalated\\\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt\", \"description\": \"Write file outside workspace and verify\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to write a file outside the workspace\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":1501,"outputTokens":174,"cacheReadTokens":0,"reasoningTokens":28}},"sourceEventSeqs":[8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130],"surfaceOp":"append"} {"type":"tool/call","seq":132,"time":1783962245385,"data":{"turn":1,"step":1,"callId":"call_00_d0sAHpJ9mYOJi0z7KNy30441","name":"bash","arguments":"{\"command\": \"printf 'escalated\\\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt\", \"description\": \"Write file outside workspace and verify\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to write a file outside the workspace\"}"}} -{"type":"approval/asked","seq":133,"time":1783962245386,"data":{"id":"bb75e939-221a-48b0-ad67-3da110ecbd31","toolName":"bash","callId":"call_00_d0sAHpJ9mYOJi0z7KNy30441","reason":"escalate sandbox to danger-full-access: the user asked to write a file outside the workspace"}} -{"type":"approval/decided","seq":134,"time":1783962245387,"data":{"id":"bb75e939-221a-48b0-ad67-3da110ecbd31","outcome":"allowed-once"}} +{"type":"approval/asked","seq":133,"time":1783962245386,"data":{"id":"21634890-d003-489f-b1b5-d38e187561c5","toolName":"bash","callId":"call_00_d0sAHpJ9mYOJi0z7KNy30441","reason":"escalate sandbox to danger-full-access: the user asked to write a file outside the workspace"}} +{"type":"approval/decided","seq":134,"time":1783962245387,"data":{"id":"21634890-d003-489f-b1b5-d38e187561c5","outcome":"allowed-once"}} {"type":"tool/result","seq":135,"time":1783962245399,"data":{"turn":1,"step":1,"callId":"call_00_d0sAHpJ9mYOJi0z7KNy30441","content":[{"type":"text","text":"escalated\n"}],"isError":false},"sourceEventSeqs":[132],"surfaceOp":"append"} {"type":"step/end","seq":136,"time":1783962245400,"data":{"turn":1,"step":1}} {"type":"step/start","seq":137,"time":1783962245400,"data":{"turn":1,"step":2}} diff --git a/examples/acp-agent/tests/snapshots/escalation-rejected/session.jsonl b/examples/acp-agent/tests/snapshots/escalation-rejected/session.jsonl index c4f5658452..0d384a55ee 100644 --- a/examples/acp-agent/tests/snapshots/escalation-rejected/session.jsonl +++ b/examples/acp-agent/tests/snapshots/escalation-rejected/session.jsonl @@ -156,8 +156,8 @@ {"type":"assistant/chunk","seq":154,"time":1783962246274,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":155,"time":1783962246274,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run a specific command with `sandbox_permissions` set to `danger-full-access` and a specific justification. They explicitly said NOT to run it without sandbox_permissions first. Let me do exactly that."},{"type":"tool-call","id":"call_00_WB1vnPomi8yr6MlcFKTj7912","name":"bash","arguments":"{\"command\": \"printf 'escalated\\\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt\", \"description\": \"Write to /tmp and verify, then clean up\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to write a file outside the workspace\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":1509,"outputTokens":198,"cacheReadTokens":0,"reasoningTokens":48}},"sourceEventSeqs":[8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154],"surfaceOp":"append"} {"type":"tool/call","seq":156,"time":1783962246274,"data":{"turn":1,"step":1,"callId":"call_00_WB1vnPomi8yr6MlcFKTj7912","name":"bash","arguments":"{\"command\": \"printf 'escalated\\\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt\", \"description\": \"Write to /tmp and verify, then clean up\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to write a file outside the workspace\"}"}} -{"type":"approval/asked","seq":157,"time":1783962246275,"data":{"id":"9a997191-81a0-4076-914b-255ebe3e6882","toolName":"bash","callId":"call_00_WB1vnPomi8yr6MlcFKTj7912","reason":"escalate sandbox to danger-full-access: the user asked to write a file outside the workspace"}} -{"type":"approval/decided","seq":158,"time":1783962246275,"data":{"id":"9a997191-81a0-4076-914b-255ebe3e6882","outcome":"rejected"}} +{"type":"approval/asked","seq":157,"time":1783962246275,"data":{"id":"4ffb95e3-8f94-41db-bfd3-2f56e6631ef8","toolName":"bash","callId":"call_00_WB1vnPomi8yr6MlcFKTj7912","reason":"escalate sandbox to danger-full-access: the user asked to write a file outside the workspace"}} +{"type":"approval/decided","seq":158,"time":1783962246275,"data":{"id":"4ffb95e3-8f94-41db-bfd3-2f56e6631ef8","outcome":"rejected"}} {"type":"tool/result","seq":159,"time":1783962246275,"data":{"turn":1,"step":1,"callId":"call_00_WB1vnPomi8yr6MlcFKTj7912","content":[{"type":"text","text":"Error: the user rejected escalating this command to \"danger-full-access\""}],"isError":true},"sourceEventSeqs":[156],"surfaceOp":"append"} {"type":"step/end","seq":160,"time":1783962246276,"data":{"turn":1,"step":1}} {"type":"step/start","seq":161,"time":1783962246276,"data":{"turn":1,"step":2}} diff --git a/examples/acp-agent/tests/snapshots/fs-escalation-approved/session.jsonl b/examples/acp-agent/tests/snapshots/fs-escalation-approved/session.jsonl index 67f3d15be0..918b3ab4d5 100644 --- a/examples/acp-agent/tests/snapshots/fs-escalation-approved/session.jsonl +++ b/examples/acp-agent/tests/snapshots/fs-escalation-approved/session.jsonl @@ -90,9 +90,9 @@ {"type":"assistant/chunk","seq":88,"time":1784045703776,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":89,"time":1784045703780,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to create a file using the write tool with sandbox_permissions. Let me do that."},{"type":"tool-call","id":"call_00_Fnymmavpr4klMDy4Fdej3227","name":"write","arguments":"{\"file_path\": \"escalated.md\", \"content\": \"escalated\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to escalate this write\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":3871,"outputTokens":132,"cacheReadTokens":0,"reasoningTokens":23}},"sourceEventSeqs":[8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88],"surfaceOp":"append"} {"type":"tool/call","seq":90,"time":1784045703780,"data":{"turn":1,"step":1,"callId":"call_00_Fnymmavpr4klMDy4Fdej3227","name":"write","arguments":"{\"file_path\": \"escalated.md\", \"content\": \"escalated\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to escalate this write\"}"}} -{"type":"approval/asked","seq":91,"time":1784045703782,"data":{"id":"f6a28337-87da-4555-bc47-2b5354300cc4","toolName":"write","callId":"call_00_Fnymmavpr4klMDy4Fdej3227","reason":"escalate sandbox to danger-full-access: the user asked to escalate this write"}} -{"type":"approval/decided","seq":92,"time":1784045703786,"data":{"id":"f6a28337-87da-4555-bc47-2b5354300cc4","outcome":"allowed-once"}} -{"type":"tool/result","seq":93,"time":1784045703798,"data":{"turn":1,"step":1,"callId":"call_00_Fnymmavpr4klMDy4Fdej3227","content":[{"type":"text","text":"/private/var/folders/2g/b32ct0qn1d728l_v6tdkjytr0000gn/T/acp-snap-cwd-vmEGzd/escalated.md\nfile\n\nCreated file\n"}],"isError":false,"meta":{"diffs":[]}},"sourceEventSeqs":[90],"surfaceOp":"append"} +{"type":"approval/asked","seq":91,"time":1784045703782,"data":{"id":"d7250dff-fc08-4163-9b51-f43f08d6f36f","toolName":"write","callId":"call_00_Fnymmavpr4klMDy4Fdej3227","reason":"escalate sandbox to danger-full-access: the user asked to escalate this write"}} +{"type":"approval/decided","seq":92,"time":1784045703786,"data":{"id":"d7250dff-fc08-4163-9b51-f43f08d6f36f","outcome":"allowed-once"}} +{"type":"tool/result","seq":93,"time":1784045703798,"data":{"turn":1,"step":1,"callId":"call_00_Fnymmavpr4klMDy4Fdej3227","content":[{"type":"text","text":"/var/folders/2g/b32ct0qn1d728l_v6tdkjytr0000gn/T/acp-snap-cwd-vmEGzd/escalated.md\nfile\n\nCreated file\n"}],"isError":false,"meta":{"diffs":[]}},"sourceEventSeqs":[90],"surfaceOp":"append"} {"type":"step/end","seq":94,"time":1784045703798,"data":{"turn":1,"step":1}} {"type":"step/start","seq":95,"time":1784045703799,"data":{"turn":1,"step":2}} {"type":"assistant/chunk","seq":96,"time":1784045704512,"data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}} diff --git a/examples/acp-agent/tests/snapshots/fs-write/session.jsonl b/examples/acp-agent/tests/snapshots/fs-write/session.jsonl index 46de5fa221..dbaaf8a8b6 100644 --- a/examples/acp-agent/tests/snapshots/fs-write/session.jsonl +++ b/examples/acp-agent/tests/snapshots/fs-write/session.jsonl @@ -63,7 +63,7 @@ {"type":"assistant/chunk","seq":61,"time":1783352079886,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":62,"time":1783352079888,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to create a file named notes.txt with the content \"hello world\" using the write tool, then reply with \"DONE\"."},{"type":"tool-call","id":"call_00_APMUCJJm9lrTSlVbg6dB0185","name":"write","arguments":"{\"file_path\": \"notes.txt\", \"content\": \"hello world\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":2891,"outputTokens":92,"cacheReadTokens":0,"reasoningTokens":30}},"sourceEventSeqs":[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],"surfaceOp":"append"} {"type":"tool/call","seq":63,"time":1783352079888,"data":{"turn":1,"step":1,"callId":"call_00_APMUCJJm9lrTSlVbg6dB0185","name":"write","arguments":"{\"file_path\": \"notes.txt\", \"content\": \"hello world\"}"}} -{"type":"tool/result","seq":64,"time":1783352079897,"data":{"turn":1,"step":1,"callId":"call_00_APMUCJJm9lrTSlVbg6dB0185","content":[{"type":"text","text":"/private/tmp/acp-snap-cwd-sNvn5N/notes.txt\nfile\n\nCreated file\n"}],"isError":false,"meta":{"diffs":[]}},"sourceEventSeqs":[63],"surfaceOp":"append"} +{"type":"tool/result","seq":64,"time":1783352079897,"data":{"turn":1,"step":1,"callId":"call_00_APMUCJJm9lrTSlVbg6dB0185","content":[{"type":"text","text":"/tmp/acp-snap-cwd-sNvn5N/notes.txt\nfile\n\nCreated file\n"}],"isError":false,"meta":{"diffs":[]}},"sourceEventSeqs":[63],"surfaceOp":"append"} {"type":"step/end","seq":65,"time":1783352079898,"data":{"turn":1,"step":1}} {"type":"step/start","seq":66,"time":1783352079899,"data":{"turn":1,"step":2}} {"type":"assistant/chunk","seq":67,"time":1783352080825,"data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}} diff --git a/examples/acp-agent/tests/snapshots/hook-cc-pretool-ask/session.jsonl b/examples/acp-agent/tests/snapshots/hook-cc-pretool-ask/session.jsonl index baa85c9d77..810083f2a9 100644 --- a/examples/acp-agent/tests/snapshots/hook-cc-pretool-ask/session.jsonl +++ b/examples/acp-agent/tests/snapshots/hook-cc-pretool-ask/session.jsonl @@ -56,8 +56,8 @@ {"type":"tool/call","seq":54,"time":1783352172557,"data":{"turn":1,"step":1,"callId":"call_00_6k0oGSliVHxGSgqBmMEO4311","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Echo HELLO\"}"}} {"type":"hook/invoked","seq":55,"time":1783352172558,"data":{"turn":1,"point":"PreToolUse","dialect":"claude","handlerId":"claude:PreToolUse:1","matcher":"bash"}} {"type":"hook/result","seq":56,"time":1783352172573,"data":{"turn":1,"point":"PreToolUse","handlerId":"claude:PreToolUse:1","decision":"ask","exitCode":0,"durationMs":14.113374999999905}} -{"type":"approval/asked","seq":57,"time":1783962235813,"data":{"id":"c7877002-3d2e-432e-9624-0b9ef3624df3","toolName":"bash","callId":"call_00_6k0oGSliVHxGSgqBmMEO4311","reason":"bash requires manual approval in this session"}} -{"type":"approval/decided","seq":58,"time":1783962235813,"data":{"id":"c7877002-3d2e-432e-9624-0b9ef3624df3","outcome":"rejected"}} +{"type":"approval/asked","seq":57,"time":1783962235813,"data":{"id":"e629ff8c-60d2-49b1-9603-333b16ef3306","toolName":"bash","callId":"call_00_6k0oGSliVHxGSgqBmMEO4311","reason":"bash requires manual approval in this session"}} +{"type":"approval/decided","seq":58,"time":1783962235813,"data":{"id":"e629ff8c-60d2-49b1-9603-333b16ef3306","outcome":"rejected"}} {"type":"tool/result","seq":59,"time":1783962235814,"data":{"turn":1,"step":1,"callId":"call_00_6k0oGSliVHxGSgqBmMEO4311","content":[{"type":"text","text":"Error: the user rejected tool \"bash\""}],"isError":true},"sourceEventSeqs":[54],"surfaceOp":"append"} {"type":"step/end","seq":60,"time":1783962235814,"data":{"turn":1,"step":1}} {"type":"step/start","seq":61,"time":1783962235814,"data":{"turn":1,"step":2}} diff --git a/examples/acp-agent/tests/snapshots/plan-mode/session.jsonl b/examples/acp-agent/tests/snapshots/plan-mode/session.jsonl index 7c1b1a7574..71a1189702 100644 --- a/examples/acp-agent/tests/snapshots/plan-mode/session.jsonl +++ b/examples/acp-agent/tests/snapshots/plan-mode/session.jsonl @@ -324,7 +324,7 @@ {"type":"assistant/chunk","seq":322,"time":1784525379646,"data":{"turn":1,"step":3,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":323,"time":1784525379647,"data":{"turn":1,"step":3,"content":[{"type":"reasoning","text":"The plan is approved. Now I need to apply exactly that one edit: change line 2 of notes.txt to say \"hello world\". I'll use the edit tool to replace the current line 2 content."},{"type":"tool-call","id":"call_00_eYuZ6fAuzqr8vtFd3oRg2452","name":"edit","arguments":"{\"file_path\": \"notes.txt\", \"old_string\": \"- the greeting message still says \\\"helo wrld\\\"\", \"new_string\": \"hello world\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":3269,"outputTokens":134,"cacheReadTokens":0,"reasoningTokens":43}},"sourceEventSeqs":[232,233,234,235,236,237,238,239,240,241,242,243,244,245,246,247,248,249,250,251,252,253,254,255,256,257,258,259,260,261,262,263,264,265,266,267,268,269,270,271,272,273,274,275,276,277,278,279,280,281,282,283,284,285,286,287,288,289,290,291,292,293,294,295,296,297,298,299,300,301,302,303,304,305,306,307,308,309,310,311,312,313,314,315,316,317,318,319,320,321,322],"surfaceOp":"append"} {"type":"tool/call","seq":324,"time":1784525379647,"data":{"turn":1,"step":3,"callId":"call_00_eYuZ6fAuzqr8vtFd3oRg2452","name":"edit","arguments":"{\"file_path\": \"notes.txt\", \"old_string\": \"- the greeting message still says \\\"helo wrld\\\"\", \"new_string\": \"hello world\"}"}} -{"type":"tool/result","seq":325,"time":1784525379652,"data":{"turn":1,"step":3,"callId":"call_00_eYuZ6fAuzqr8vtFd3oRg2452","content":[{"type":"text","text":"Error: edit requires reading \"/private/var/folders/2g/b32ct0qn1d728l_v6tdkjytr0000gn/T/acp-snap-cwd-tia05W/notes.txt\" first"}],"isError":true,"error":{"name":"FsError","code":"FS_NOT_OBSERVED"}},"sourceEventSeqs":[324],"surfaceOp":"append"} +{"type":"tool/result","seq":325,"time":1784525379652,"data":{"turn":1,"step":3,"callId":"call_00_eYuZ6fAuzqr8vtFd3oRg2452","content":[{"type":"text","text":"Error: edit requires reading \"/var/folders/2g/b32ct0qn1d728l_v6tdkjytr0000gn/T/acp-snap-cwd-tia05W/notes.txt\" first"}],"isError":true,"error":{"name":"FsError","code":"FS_NOT_OBSERVED"}},"sourceEventSeqs":[324],"surfaceOp":"append"} {"type":"step/end","seq":326,"time":1784525379652,"data":{"turn":1,"step":3}} {"type":"step/start","seq":327,"time":1784525379655,"data":{"turn":1,"step":4}} {"type":"assistant/chunk","seq":328,"time":1784525380205,"data":{"turn":1,"step":4,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}} @@ -428,7 +428,7 @@ {"type":"assistant/chunk","seq":426,"time":1784525381745,"data":{"turn":1,"step":5,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":427,"time":1784525381746,"data":{"turn":1,"step":5,"content":[{"type":"reasoning","text":"Now I can apply the edit."},{"type":"tool-call","id":"call_00_7H6VFdLjBqzzW9M5Z4OR8016","name":"edit","arguments":"{\"file_path\": \"notes.txt\", \"old_string\": \"- the greeting message still says \\\"helo wrld\\\"\", \"new_string\": \"hello world\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":169,"outputTokens":98,"cacheReadTokens":3456,"reasoningTokens":7}},"sourceEventSeqs":[372,373,374,375,376,377,378,379,380,381,382,383,384,385,386,387,388,389,390,391,392,393,394,395,396,397,398,399,400,401,402,403,404,405,406,407,408,409,410,411,412,413,414,415,416,417,418,419,420,421,422,423,424,425,426],"surfaceOp":"append"} {"type":"tool/call","seq":428,"time":1784525381746,"data":{"turn":1,"step":5,"callId":"call_00_7H6VFdLjBqzzW9M5Z4OR8016","name":"edit","arguments":"{\"file_path\": \"notes.txt\", \"old_string\": \"- the greeting message still says \\\"helo wrld\\\"\", \"new_string\": \"hello world\"}"}} -{"type":"tool/result","seq":429,"time":1784525381766,"data":{"turn":1,"step":5,"callId":"call_00_7H6VFdLjBqzzW9M5Z4OR8016","content":[{"type":"text","text":"The file /private/var/folders/2g/b32ct0qn1d728l_v6tdkjytr0000gn/T/acp-snap-cwd-tia05W/notes.txt has been updated successfully."}],"isError":false,"meta":{"diffs":[{"path":"notes.txt","oldText":"project notes\n- the greeting message still says \"helo wrld\"","newText":"project notes\nhello world"}]}},"sourceEventSeqs":[428],"surfaceOp":"append"} +{"type":"tool/result","seq":429,"time":1784525381766,"data":{"turn":1,"step":5,"callId":"call_00_7H6VFdLjBqzzW9M5Z4OR8016","content":[{"type":"text","text":"The file /var/folders/2g/b32ct0qn1d728l_v6tdkjytr0000gn/T/acp-snap-cwd-tia05W/notes.txt has been updated successfully."}],"isError":false,"meta":{"diffs":[{"path":"notes.txt","oldText":"project notes\n- the greeting message still says \"helo wrld\"","newText":"project notes\nhello world"}]}},"sourceEventSeqs":[428],"surfaceOp":"append"} {"type":"step/end","seq":430,"time":1784525381766,"data":{"turn":1,"step":5}} {"type":"step/start","seq":431,"time":1784525381768,"data":{"turn":1,"step":6}} {"type":"assistant/chunk","seq":432,"time":1784525382334,"data":{"turn":1,"step":6,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}} diff --git a/packages/goal/tool-goal/tests/tool-goal.spec.ts b/packages/goal/tool-goal/tests/tool-goal.spec.ts index c743f84903..30270fd34d 100644 --- a/packages/goal/tool-goal/tests/tool-goal.spec.ts +++ b/packages/goal/tool-goal/tests/tool-goal.spec.ts @@ -409,18 +409,10 @@ describe('goal tool state transitions', () => { } const paused = await execute(ctx, 'update_goal', { - goal_id: goal.id, - revision: goal.revision, - action: 'pause', - objective: 'ignored replacement', - max_goal_rounds: 1, - blocked_reason: 'ignored blocker', + goal_id: goal.id, revision: goal.revision, action: 'pause', ...strictArgs, }, root.agent) goal = ctx.goals.get(root.agent)! - expect(resultGoal(paused)).toMatchObject({ - phase: 'paused', objective: 'valid', maxGoalRounds: goal.maxGoalRounds, - }) - expect(goal.blockedReason).toBeUndefined() + expect(resultGoal(paused)).toMatchObject({ phase: 'paused', objective: 'valid' }) const resumed = await execute(ctx, 'update_goal', { goal_id: goal.id, revision: goal.revision, action: 'resume', ...strictArgs, }, root.agent) From 0d413490021936d640320f076af0bfef248fa949 Mon Sep 17 00:00:00 2001 From: Turtle Date: Fri, 24 Jul 2026 13:09:13 +0800 Subject: [PATCH 4/8] Revert "fix(goal): tolerate strict provider filler fields" This reverts commit 9a30dbb0f537f512180754c4778c15f9f94f1a95. --- ...26-07-19-model-facing-goal-tools.i18n.yaml | 4 +- .../2026-07-19-model-facing-goal-tools.md | 4 +- .../2026-07-19-model-facing-goal-tools.zh.md | 4 +- .../tool-schemas.expected.json | 8 +-- .../both-mode-turn/tool-schemas.expected.json | 8 +-- .../escalation-approved/session.jsonl | 4 +- .../escalation-rejected/session.jsonl | 4 +- .../tests/snapshots/fs-edit/session.jsonl | 2 +- .../snapshots/fs-policy-reject/session.jsonl | 4 +- .../fs-write-overwrite/session.jsonl | 2 +- .../hook-cc-pretool-ask/session.jsonl | 4 +- .../lsp-definition/tool-schemas.expected.json | 8 +-- .../tool-schemas.expected.json | 16 ++--- .../tool-schemas.expected.json | 16 ++--- .../plan-mode/tool-schemas.expected.json | 16 ++--- .../pty-tools/tool-schemas.expected.json | 8 +-- .../skill-load/tool-schemas.expected.json | 8 +-- .../text-turn/tool-schemas.expected.json | 8 +-- .../tool-schemas.expected.json | 8 +-- packages/goal/tool-goal/README.md | 2 +- packages/goal/tool-goal/src/index.ts | 29 +++++++-- .../goal/tool-goal/tests/tool-goal.spec.ts | 61 ++++++++----------- 22 files changed, 119 insertions(+), 109 deletions(-) diff --git a/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.i18n.yaml b/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.i18n.yaml index b91a8758ca..e53c591aa5 100644 --- a/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.i18n.yaml +++ b/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.i18n.yaml @@ -2,5 +2,5 @@ # side as of the last confirmed-consistent state. Both languages carry equal authority; # after editing either side, bring the other along and re-record with: # pnpm run verify-translation-pairing --write -2026-07-19-model-facing-goal-tools.md: 7df04678ab8f78be71504b9a0d6e3a05391833e3 -2026-07-19-model-facing-goal-tools.zh.md: 102337fd57291102ce16a3978bafe9fcf62b9695 +2026-07-19-model-facing-goal-tools.md: 7cc3907d708115207e166455ea988120a03d768b +2026-07-19-model-facing-goal-tools.zh.md: 1a381160354d6a2a24f957f41bc9e375c1ab01ca diff --git a/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.md b/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.md index 7df04678ab..7cc3907d70 100644 --- a/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.md +++ b/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.md @@ -16,7 +16,7 @@ The surface also needs to preserve the separation between durable state and live ### Tools and model contract -`get_goal()` returns the current goal or `null`. A non-null result contains the compare-and-set id and revision, objective, durable phase, admitted and maximum goal rounds, any blocker reason, plus the process-local activation observation. `create_goal(objective, max_goal_rounds?)` creates one long-running same-session objective. `update_goal(goal_id, revision, action, objective?, max_goal_rounds?, blocked_reason?)` supports `edit`, `pause`, `resume`, `complete`, and `blocked`; each action consumes only its fields and ignores the others, while `blocked` still requires a non-empty `blocked_reason` and persists it under the stable `model-reported` code. This tolerance is necessary because strict-schema providers may populate every declared property even though JSON Schema marks the action-specific fields optional. +`get_goal()` returns the current goal or `null`. A non-null result contains the compare-and-set id and revision, objective, durable phase, admitted and maximum goal rounds, any blocker reason, plus the process-local activation observation. `create_goal(objective, max_goal_rounds?)` creates one long-running same-session objective. `update_goal(goal_id, revision, action, objective?, max_goal_rounds?, blocked_reason?)` supports `edit`, `pause`, `resume`, `complete`, and `blocked`; replacement fields are valid only for `edit`, while a non-empty `blocked_reason` is required only for `blocked` and persists under the stable `model-reported` code. The prompt tells the model that it may infer goal intent from a direct human request in any wording or language, but should not convert routine single-turn work into a goal. It must read the current goal before updating and copy the exact id and revision. On a restored or forked active-but-disarmed goal, a semantic human request to continue is grounds for `resume`. Completion is reserved for an achieved objective, and difficulty or uncertainty alone is not a blocker; a block report must name the concrete condition. @@ -38,7 +38,7 @@ Complete and blocked accept either direct-human authority or the exact current g ## Testing -Unit coverage pins registration and disposal, exclusive scheduling, generated prompt policy, generic presentation, direct-human creation in a non-English turn, exact/stale/non-running agent and driver checks, live-child rejection, resumed-fork root authority, steering, mismatched initiators, read/create/edit/pause/resume behavior, required blocker explanations, strict-provider filler fields, rearming after a session-start edge, authority-before-argument failures, exact goal-round completion, autonomous-only terminal stopping, the configured blocking threshold, and immediate human blocking. A keyless replay snapshot mounts the goal domain and tools into the real headless one-shot application, drives `create_goal` and `get_goal` through the shipped loop and persistence stack, pins its stream-json transcript, and inspects the externally persisted goal change. The echo-agent fixture is intentionally not used as an application-UX surrogate. +Unit coverage pins registration and disposal, exclusive scheduling, generated prompt policy, generic presentation, direct-human creation in a non-English turn, exact/stale/non-running agent and driver checks, live-child rejection, resumed-fork root authority, steering, mismatched initiators, read/create/edit/pause/resume behavior, conditional blocker explanations, rearming after a session-start edge, authority-before-conditional-argument failures, exact goal-round completion, autonomous-only terminal stopping, the configured blocking threshold, and immediate human blocking. A keyless replay snapshot mounts the goal domain and tools into the real headless one-shot application, drives `create_goal` and `get_goal` through the shipped loop and persistence stack, pins its stream-json transcript, and inspects the externally persisted goal change. The echo-agent fixture is intentionally not used as an application-UX surrogate. ## Alternatives considered diff --git a/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.zh.md b/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.zh.md index 102337fd57..1a38116035 100644 --- a/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.zh.md +++ b/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.zh.md @@ -16,7 +16,7 @@ Status: implemented ### 工具与模型契约 -`get_goal()` 返回当前目标或 `null`。非空结果包含用于比较并交换的 id 与修订号、目标描述、持久阶段、已接纳和最大目标回合数、可能存在的阻塞原因，以及进程本地激活态观察。`create_goal(objective, max_goal_rounds?)` 创建一个长时间运行的同会话目标。`update_goal(goal_id, revision, action, objective?, max_goal_rounds?, blocked_reason?)` 支持 `edit`、`pause`、`resume`、`complete` 和 `blocked`；每个动作只消费自己的字段并忽略其他字段，而 `blocked` 仍要求非空的 `blocked_reason`，并以稳定代码 `model-reported` 持久化。尽管 JSON Schema 将动作专用字段标记为可选，严格模式提供方仍可能填充所有已声明属性，因此需要这种容忍行为。 +`get_goal()` 返回当前目标或 `null`。非空结果包含用于比较并交换的 id 与修订号、目标描述、持久阶段、已接纳和最大目标回合数、可能存在的阻塞原因，以及进程本地激活态观察。`create_goal(objective, max_goal_rounds?)` 创建一个长时间运行的同会话目标。`update_goal(goal_id, revision, action, objective?, max_goal_rounds?, blocked_reason?)` 支持 `edit`、`pause`、`resume`、`complete` 和 `blocked`；替换字段仅对 `edit` 有效，非空的 `blocked_reason` 仅在 `blocked` 时必填，并以稳定代码 `model-reported` 持久化。 提示词告诉模型：它可以从任何措辞或语言的直接人类请求中推断目标意图，但不应把常规单轮工作转换为目标。更新前必须读取当前目标，并复制准确的 id 和修订号。对于恢复或派生后处于活跃但未激活状态的目标，人类在语义上要求继续即可成为执行 `resume` 的依据。只有目标已经实现时才能标记完成，困难或不确定性本身不构成阻塞；阻塞报告必须说明具体条件。 @@ -38,7 +38,7 @@ Status: implemented ## 测试 -单元测试固定注册与释放、独占调度、生成的提示词策略、通用展示、非英语轮次中的直接人类创建、精确/陈旧/非运行中智能体与驱动检查、实时子智能体拒绝、恢复后派生根的权限、steering、发起者不匹配、读取/创建/编辑/暂停/恢复行为、必需的阻塞说明、严格模式提供方填充字段、会话启动边沿后的重新激活、权限先于参数失败、准确目标回合的完成、仅自主回合触发终止、可配置阻塞阈值，以及人类立即阻塞。无密钥回放快照把目标领域和工具挂载到真实的 headless 单次运行应用中，通过随附循环与持久化栈驱动 `create_goal` 和 `get_goal`，固定 stream-json 转录，并检查外部持久化的目标变更。这里有意不把 echo-agent 测试夹具当作应用 UX 的替代品。 +单元测试固定注册与释放、独占调度、生成的提示词策略、通用展示、非英语轮次中的直接人类创建、精确/陈旧/非运行中智能体与驱动检查、实时子智能体拒绝、恢复后派生根的权限、steering、发起者不匹配、读取/创建/编辑/暂停/恢复行为、条件式阻塞说明、会话启动边沿后的重新激活、权限先于条件参数失败、准确目标回合的完成、仅自主回合触发终止、可配置阻塞阈值，以及人类立即阻塞。无密钥回放快照把目标领域和工具挂载到真实的 headless 单次运行应用中，通过随附循环与持久化栈驱动 `create_goal` 和 `get_goal`，固定 stream-json 转录，并检查外部持久化的目标变更。这里有意不把 echo-agent 测试夹具当作应用 UX 的替代品。 ## 考虑过的替代方案 diff --git a/examples/acp-agent/tests/snapshots/advanced-toolchain/tool-schemas.expected.json b/examples/acp-agent/tests/snapshots/advanced-toolchain/tool-schemas.expected.json index b3f147fdbf..6b50a5d220 100644 --- a/examples/acp-agent/tests/snapshots/advanced-toolchain/tool-schemas.expected.json +++ b/examples/acp-agent/tests/snapshots/advanced-toolchain/tool-schemas.expected.json @@ -470,7 +470,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -495,15 +495,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective for action edit; ignored for other actions." + "description": "Replacement objective; valid only with action edit." }, "max_goal_rounds": { "type": "number", - "description": "Replacement round cap for action edit; ignored for other actions." + "description": "Replacement cap; valid only with action edit." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." + "description": "Concrete blocking condition; required only with action blocked." } }, "required": [ diff --git a/examples/acp-agent/tests/snapshots/both-mode-turn/tool-schemas.expected.json b/examples/acp-agent/tests/snapshots/both-mode-turn/tool-schemas.expected.json index 0e2ebef051..7ceeec4042 100644 --- a/examples/acp-agent/tests/snapshots/both-mode-turn/tool-schemas.expected.json +++ b/examples/acp-agent/tests/snapshots/both-mode-turn/tool-schemas.expected.json @@ -413,7 +413,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -438,15 +438,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective for action edit; ignored for other actions." + "description": "Replacement objective; valid only with action edit." }, "max_goal_rounds": { "type": "number", - "description": "Replacement round cap for action edit; ignored for other actions." + "description": "Replacement cap; valid only with action edit." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." + "description": "Concrete blocking condition; required only with action blocked." } }, "required": [ diff --git a/examples/acp-agent/tests/snapshots/escalation-approved/session.jsonl b/examples/acp-agent/tests/snapshots/escalation-approved/session.jsonl index 6aa7fda0a8..4438550b97 100644 --- a/examples/acp-agent/tests/snapshots/escalation-approved/session.jsonl +++ b/examples/acp-agent/tests/snapshots/escalation-approved/session.jsonl @@ -132,8 +132,8 @@ {"type":"assistant/chunk","seq":130,"time":1783962245385,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":131,"time":1783962245385,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run a command with sandbox_permissions set to danger-full-access, no prior run needed, justified as instructed."},{"type":"tool-call","id":"call_00_d0sAHpJ9mYOJi0z7KNy30441","name":"bash","arguments":"{\"command\": \"printf 'escalated\\\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt\", \"description\": \"Write file outside workspace and verify\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to write a file outside the workspace\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":1501,"outputTokens":174,"cacheReadTokens":0,"reasoningTokens":28}},"sourceEventSeqs":[8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130],"surfaceOp":"append"} {"type":"tool/call","seq":132,"time":1783962245385,"data":{"turn":1,"step":1,"callId":"call_00_d0sAHpJ9mYOJi0z7KNy30441","name":"bash","arguments":"{\"command\": \"printf 'escalated\\\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt\", \"description\": \"Write file outside workspace and verify\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to write a file outside the workspace\"}"}} -{"type":"approval/asked","seq":133,"time":1783962245386,"data":{"id":"21634890-d003-489f-b1b5-d38e187561c5","toolName":"bash","callId":"call_00_d0sAHpJ9mYOJi0z7KNy30441","reason":"escalate sandbox to danger-full-access: the user asked to write a file outside the workspace"}} -{"type":"approval/decided","seq":134,"time":1783962245387,"data":{"id":"21634890-d003-489f-b1b5-d38e187561c5","outcome":"allowed-once"}} +{"type":"approval/asked","seq":133,"time":1783962245386,"data":{"id":"efa76e4a-8e3e-4be0-9608-a68def7354b9","toolName":"bash","callId":"call_00_d0sAHpJ9mYOJi0z7KNy30441","reason":"escalate sandbox to danger-full-access: the user asked to write a file outside the workspace"}} +{"type":"approval/decided","seq":134,"time":1783962245387,"data":{"id":"efa76e4a-8e3e-4be0-9608-a68def7354b9","outcome":"allowed-once"}} {"type":"tool/result","seq":135,"time":1783962245399,"data":{"turn":1,"step":1,"callId":"call_00_d0sAHpJ9mYOJi0z7KNy30441","content":[{"type":"text","text":"escalated\n"}],"isError":false},"sourceEventSeqs":[132],"surfaceOp":"append"} {"type":"step/end","seq":136,"time":1783962245400,"data":{"turn":1,"step":1}} {"type":"step/start","seq":137,"time":1783962245400,"data":{"turn":1,"step":2}} diff --git a/examples/acp-agent/tests/snapshots/escalation-rejected/session.jsonl b/examples/acp-agent/tests/snapshots/escalation-rejected/session.jsonl index 0d384a55ee..eb01e7443e 100644 --- a/examples/acp-agent/tests/snapshots/escalation-rejected/session.jsonl +++ b/examples/acp-agent/tests/snapshots/escalation-rejected/session.jsonl @@ -156,8 +156,8 @@ {"type":"assistant/chunk","seq":154,"time":1783962246274,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":155,"time":1783962246274,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to run a specific command with `sandbox_permissions` set to `danger-full-access` and a specific justification. They explicitly said NOT to run it without sandbox_permissions first. Let me do exactly that."},{"type":"tool-call","id":"call_00_WB1vnPomi8yr6MlcFKTj7912","name":"bash","arguments":"{\"command\": \"printf 'escalated\\\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt\", \"description\": \"Write to /tmp and verify, then clean up\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to write a file outside the workspace\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":1509,"outputTokens":198,"cacheReadTokens":0,"reasoningTokens":48}},"sourceEventSeqs":[8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154],"surfaceOp":"append"} {"type":"tool/call","seq":156,"time":1783962246274,"data":{"turn":1,"step":1,"callId":"call_00_WB1vnPomi8yr6MlcFKTj7912","name":"bash","arguments":"{\"command\": \"printf 'escalated\\\\n' > /tmp/dsh-escalated.txt && cat /tmp/dsh-escalated.txt && rm /tmp/dsh-escalated.txt\", \"description\": \"Write to /tmp and verify, then clean up\", \"sandbox_permissions\": \"danger-full-access\", \"justification\": \"the user asked to write a file outside the workspace\"}"}} -{"type":"approval/asked","seq":157,"time":1783962246275,"data":{"id":"4ffb95e3-8f94-41db-bfd3-2f56e6631ef8","toolName":"bash","callId":"call_00_WB1vnPomi8yr6MlcFKTj7912","reason":"escalate sandbox to danger-full-access: the user asked to write a file outside the workspace"}} -{"type":"approval/decided","seq":158,"time":1783962246275,"data":{"id":"4ffb95e3-8f94-41db-bfd3-2f56e6631ef8","outcome":"rejected"}} +{"type":"approval/asked","seq":157,"time":1783962246275,"data":{"id":"e5cc84ef-8704-4cd9-8e47-897a0a51beb6","toolName":"bash","callId":"call_00_WB1vnPomi8yr6MlcFKTj7912","reason":"escalate sandbox to danger-full-access: the user asked to write a file outside the workspace"}} +{"type":"approval/decided","seq":158,"time":1783962246275,"data":{"id":"e5cc84ef-8704-4cd9-8e47-897a0a51beb6","outcome":"rejected"}} {"type":"tool/result","seq":159,"time":1783962246275,"data":{"turn":1,"step":1,"callId":"call_00_WB1vnPomi8yr6MlcFKTj7912","content":[{"type":"text","text":"Error: the user rejected escalating this command to \"danger-full-access\""}],"isError":true},"sourceEventSeqs":[156],"surfaceOp":"append"} {"type":"step/end","seq":160,"time":1783962246276,"data":{"turn":1,"step":1}} {"type":"step/start","seq":161,"time":1783962246276,"data":{"turn":1,"step":2}} diff --git a/examples/acp-agent/tests/snapshots/fs-edit/session.jsonl b/examples/acp-agent/tests/snapshots/fs-edit/session.jsonl index bcdd4c7b99..d0bd2d5bef 100644 --- a/examples/acp-agent/tests/snapshots/fs-edit/session.jsonl +++ b/examples/acp-agent/tests/snapshots/fs-edit/session.jsonl @@ -130,7 +130,7 @@ {"type":"assistant/chunk","seq":128,"time":1783352087469,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":129,"time":1783352087469,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"Now I need to replace \"DEBUG\" with \"RELEASE\" using the edit tool."},{"type":"tool-call","id":"call_00_vOytneZ0XpsLslEEJAxR6398","name":"edit","arguments":"{\"file_path\": \"config.txt\", \"old_string\": \"DEBUG\", \"new_string\": \"RELEASE\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":241,"outputTokens":98,"cacheReadTokens":2816,"reasoningTokens":18}},"sourceEventSeqs":[74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128],"surfaceOp":"append"} {"type":"tool/call","seq":130,"time":1783352087469,"data":{"turn":1,"step":2,"callId":"call_00_vOytneZ0XpsLslEEJAxR6398","name":"edit","arguments":"{\"file_path\": \"config.txt\", \"old_string\": \"DEBUG\", \"new_string\": \"RELEASE\"}"}} -{"type":"tool/result","seq":131,"time":1783352087476,"data":{"turn":1,"step":2,"callId":"call_00_vOytneZ0XpsLslEEJAxR6398","content":[{"type":"text","text":"The file /private/tmp/acp-snap-cwd-0BxHdV/config.txt has been updated successfully."}],"isError":false,"meta":{"diffs":[{"path":"config.txt","oldText":"mode=DEBUG\nlevel=info","newText":"mode=RELEASE\nlevel=info"}]}},"sourceEventSeqs":[130],"surfaceOp":"append"} +{"type":"tool/result","seq":131,"time":1783352087476,"data":{"turn":1,"step":2,"callId":"call_00_vOytneZ0XpsLslEEJAxR6398","content":[{"type":"text","text":"The file /tmp/acp-snap-cwd-0BxHdV/config.txt has been updated successfully."}],"isError":false,"meta":{"diffs":[{"path":"config.txt","oldText":"mode=DEBUG\nlevel=info","newText":"mode=RELEASE\nlevel=info"}]}},"sourceEventSeqs":[130],"surfaceOp":"append"} {"type":"step/end","seq":132,"time":1783352087477,"data":{"turn":1,"step":2}} {"type":"step/start","seq":133,"time":1783352087477,"data":{"turn":1,"step":3}} {"type":"assistant/chunk","seq":134,"time":1783352088286,"data":{"turn":1,"step":3,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}} diff --git a/examples/acp-agent/tests/snapshots/fs-policy-reject/session.jsonl b/examples/acp-agent/tests/snapshots/fs-policy-reject/session.jsonl index 25c57ee964..8f97feff04 100644 --- a/examples/acp-agent/tests/snapshots/fs-policy-reject/session.jsonl +++ b/examples/acp-agent/tests/snapshots/fs-policy-reject/session.jsonl @@ -78,7 +78,7 @@ {"type":"assistant/chunk","seq":76,"time":1783611703969,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":77,"time":1783611703972,"data":{"turn":1,"step":1,"content":[{"type":"reasoning","text":"The user wants me to use the edit tool to replace \"blue\" with \"green\" in settings.txt without reading the file first, and then reply with just \"DONE\"."},{"type":"tool-call","id":"call_00_x0zlnXl5JOxLrAYL9y7P0119","name":"edit","arguments":"{\"file_path\": \"settings.txt\", \"old_string\": \"blue\", \"new_string\": \"green\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":3132,"outputTokens":115,"cacheReadTokens":0,"reasoningTokens":36}},"sourceEventSeqs":[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],"surfaceOp":"append"} {"type":"tool/call","seq":78,"time":1783611703972,"data":{"turn":1,"step":1,"callId":"call_00_x0zlnXl5JOxLrAYL9y7P0119","name":"edit","arguments":"{\"file_path\": \"settings.txt\", \"old_string\": \"blue\", \"new_string\": \"green\"}"}} -{"type":"tool/result","seq":79,"time":1783611703978,"data":{"turn":1,"step":1,"callId":"call_00_x0zlnXl5JOxLrAYL9y7P0119","content":[{"type":"text","text":"Error: edit requires reading \"/private/var/folders/2c/psb0_fmx7hbgz558xjt_f0l00000gn/T/acp-snap-cwd-QzoqnB/settings.txt\" first"}],"isError":true,"error":{"name":"FsError","code":"FS_NOT_OBSERVED"}},"sourceEventSeqs":[78],"surfaceOp":"append"} +{"type":"tool/result","seq":79,"time":1783611703978,"data":{"turn":1,"step":1,"callId":"call_00_x0zlnXl5JOxLrAYL9y7P0119","content":[{"type":"text","text":"Error: edit requires reading \"/var/folders/2c/psb0_fmx7hbgz558xjt_f0l00000gn/T/acp-snap-cwd-QzoqnB/settings.txt\" first"}],"isError":true,"error":{"name":"FsError","code":"FS_NOT_OBSERVED"}},"sourceEventSeqs":[78],"surfaceOp":"append"} {"type":"step/end","seq":80,"time":1783611703978,"data":{"turn":1,"step":1}} {"type":"step/start","seq":81,"time":1783611703978,"data":{"turn":1,"step":2}} {"type":"assistant/chunk","seq":82,"time":1783611704825,"data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}} @@ -226,7 +226,7 @@ {"type":"assistant/chunk","seq":224,"time":1783611707096,"data":{"turn":1,"step":3,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":225,"time":1783611707097,"data":{"turn":1,"step":3,"content":[{"type":"reasoning","text":"The file contains \"color: blue\". I need to replace \"blue\" with \"green\". The edit tool said it requires reading first — now I've read it, so the edit should work."},{"type":"tool-call","id":"call_00_GVknJu2tksKkP4lALCwh0926","name":"edit","arguments":"{\"file_path\": \"settings.txt\", \"old_string\": \"blue\", \"new_string\": \"green\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":281,"outputTokens":119,"cacheReadTokens":3200,"reasoningTokens":40}},"sourceEventSeqs":[149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188,189,190,191,192,193,194,195,196,197,198,199,200,201,202,203,204,205,206,207,208,209,210,211,212,213,214,215,216,217,218,219,220,221,222,223,224],"surfaceOp":"append"} {"type":"tool/call","seq":226,"time":1783611707097,"data":{"turn":1,"step":3,"callId":"call_00_GVknJu2tksKkP4lALCwh0926","name":"edit","arguments":"{\"file_path\": \"settings.txt\", \"old_string\": \"blue\", \"new_string\": \"green\"}"}} -{"type":"tool/result","seq":227,"time":1783611707114,"data":{"turn":1,"step":3,"callId":"call_00_GVknJu2tksKkP4lALCwh0926","content":[{"type":"text","text":"The file /private/var/folders/2c/psb0_fmx7hbgz558xjt_f0l00000gn/T/acp-snap-cwd-QzoqnB/settings.txt has been updated successfully."}],"isError":false,"meta":{"diffs":[{"path":"settings.txt","oldText":"color: blue","newText":"color: green"}]}},"sourceEventSeqs":[226],"surfaceOp":"append"} +{"type":"tool/result","seq":227,"time":1783611707114,"data":{"turn":1,"step":3,"callId":"call_00_GVknJu2tksKkP4lALCwh0926","content":[{"type":"text","text":"The file /var/folders/2c/psb0_fmx7hbgz558xjt_f0l00000gn/T/acp-snap-cwd-QzoqnB/settings.txt has been updated successfully."}],"isError":false,"meta":{"diffs":[{"path":"settings.txt","oldText":"color: blue","newText":"color: green"}]}},"sourceEventSeqs":[226],"surfaceOp":"append"} {"type":"step/end","seq":228,"time":1783611707114,"data":{"turn":1,"step":3}} {"type":"step/start","seq":229,"time":1783611707114,"data":{"turn":1,"step":4}} {"type":"assistant/chunk","seq":230,"time":1783611707747,"data":{"turn":1,"step":4,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}} diff --git a/examples/acp-agent/tests/snapshots/fs-write-overwrite/session.jsonl b/examples/acp-agent/tests/snapshots/fs-write-overwrite/session.jsonl index f3db3493b5..5a3fc5696b 100644 --- a/examples/acp-agent/tests/snapshots/fs-write-overwrite/session.jsonl +++ b/examples/acp-agent/tests/snapshots/fs-write-overwrite/session.jsonl @@ -115,7 +115,7 @@ {"type":"assistant/chunk","seq":113,"time":1783352094988,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}} {"type":"assistant/message","seq":114,"time":1783352094988,"data":{"turn":1,"step":2,"content":[{"type":"reasoning","text":"The file contains \"original contents\". Now I'll replace it with \"replaced\"."},{"type":"tool-call","id":"call_00_N23EvXjDo4c8enyWpIUq4043","name":"write","arguments":"{\"file_path\": \"data.txt\", \"content\": \"replaced\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":228,"outputTokens":79,"cacheReadTokens":2816,"reasoningTokens":17}},"sourceEventSeqs":[70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113],"surfaceOp":"append"} {"type":"tool/call","seq":115,"time":1783352094988,"data":{"turn":1,"step":2,"callId":"call_00_N23EvXjDo4c8enyWpIUq4043","name":"write","arguments":"{\"file_path\": \"data.txt\", \"content\": \"replaced\"}"}} -{"type":"tool/result","seq":116,"time":1783352094995,"data":{"turn":1,"step":2,"callId":"call_00_N23EvXjDo4c8enyWpIUq4043","content":[{"type":"text","text":"/private/tmp/acp-snap-cwd-hH2sGY/data.txt\nfile\n\nUpdated file\n"}],"isError":false,"meta":{"diffs":[{"path":"data.txt","oldText":"original contents","newText":"replaced"}]}},"sourceEventSeqs":[115],"surfaceOp":"append"} +{"type":"tool/result","seq":116,"time":1783352094995,"data":{"turn":1,"step":2,"callId":"call_00_N23EvXjDo4c8enyWpIUq4043","content":[{"type":"text","text":"/tmp/acp-snap-cwd-hH2sGY/data.txt\nfile\n\nUpdated file\n"}],"isError":false,"meta":{"diffs":[{"path":"data.txt","oldText":"original contents","newText":"replaced"}]}},"sourceEventSeqs":[115],"surfaceOp":"append"} {"type":"step/end","seq":117,"time":1783352094995,"data":{"turn":1,"step":2}} {"type":"step/start","seq":118,"time":1783352094995,"data":{"turn":1,"step":3}} {"type":"assistant/chunk","seq":119,"time":1783352096090,"data":{"turn":1,"step":3,"chunk":{"type":"block-start","index":0,"blockType":"reasoning"}}} diff --git a/examples/acp-agent/tests/snapshots/hook-cc-pretool-ask/session.jsonl b/examples/acp-agent/tests/snapshots/hook-cc-pretool-ask/session.jsonl index 810083f2a9..888f2f5c13 100644 --- a/examples/acp-agent/tests/snapshots/hook-cc-pretool-ask/session.jsonl +++ b/examples/acp-agent/tests/snapshots/hook-cc-pretool-ask/session.jsonl @@ -56,8 +56,8 @@ {"type":"tool/call","seq":54,"time":1783352172557,"data":{"turn":1,"step":1,"callId":"call_00_6k0oGSliVHxGSgqBmMEO4311","name":"bash","arguments":"{\"command\": \"echo HELLO\", \"description\": \"Echo HELLO\"}"}} {"type":"hook/invoked","seq":55,"time":1783352172558,"data":{"turn":1,"point":"PreToolUse","dialect":"claude","handlerId":"claude:PreToolUse:1","matcher":"bash"}} {"type":"hook/result","seq":56,"time":1783352172573,"data":{"turn":1,"point":"PreToolUse","handlerId":"claude:PreToolUse:1","decision":"ask","exitCode":0,"durationMs":14.113374999999905}} -{"type":"approval/asked","seq":57,"time":1783962235813,"data":{"id":"e629ff8c-60d2-49b1-9603-333b16ef3306","toolName":"bash","callId":"call_00_6k0oGSliVHxGSgqBmMEO4311","reason":"bash requires manual approval in this session"}} -{"type":"approval/decided","seq":58,"time":1783962235813,"data":{"id":"e629ff8c-60d2-49b1-9603-333b16ef3306","outcome":"rejected"}} +{"type":"approval/asked","seq":57,"time":1783962235813,"data":{"id":"97616288-1a5e-4110-a75d-7616a24adcc4","toolName":"bash","callId":"call_00_6k0oGSliVHxGSgqBmMEO4311","reason":"bash requires manual approval in this session"}} +{"type":"approval/decided","seq":58,"time":1783962235813,"data":{"id":"97616288-1a5e-4110-a75d-7616a24adcc4","outcome":"rejected"}} {"type":"tool/result","seq":59,"time":1783962235814,"data":{"turn":1,"step":1,"callId":"call_00_6k0oGSliVHxGSgqBmMEO4311","content":[{"type":"text","text":"Error: the user rejected tool \"bash\""}],"isError":true},"sourceEventSeqs":[54],"surfaceOp":"append"} {"type":"step/end","seq":60,"time":1783962235814,"data":{"turn":1,"step":1}} {"type":"step/start","seq":61,"time":1783962235814,"data":{"turn":1,"step":2}} diff --git a/examples/acp-agent/tests/snapshots/lsp-definition/tool-schemas.expected.json b/examples/acp-agent/tests/snapshots/lsp-definition/tool-schemas.expected.json index 8359d03017..b42a434388 100644 --- a/examples/acp-agent/tests/snapshots/lsp-definition/tool-schemas.expected.json +++ b/examples/acp-agent/tests/snapshots/lsp-definition/tool-schemas.expected.json @@ -434,7 +434,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -459,15 +459,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective for action edit; ignored for other actions." + "description": "Replacement objective; valid only with action edit." }, "max_goal_rounds": { "type": "number", - "description": "Replacement round cap for action edit; ignored for other actions." + "description": "Replacement cap; valid only with action edit." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." + "description": "Concrete blocking condition; required only with action blocked." } }, "required": [ diff --git a/examples/acp-agent/tests/snapshots/model-switching/tool-schemas.expected.json b/examples/acp-agent/tests/snapshots/model-switching/tool-schemas.expected.json index dac439b700..ef40784fa5 100644 --- a/examples/acp-agent/tests/snapshots/model-switching/tool-schemas.expected.json +++ b/examples/acp-agent/tests/snapshots/model-switching/tool-schemas.expected.json @@ -397,7 +397,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -422,15 +422,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective for action edit; ignored for other actions." + "description": "Replacement objective; valid only with action edit." }, "max_goal_rounds": { "type": "number", - "description": "Replacement round cap for action edit; ignored for other actions." + "description": "Replacement cap; valid only with action edit." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." + "description": "Concrete blocking condition; required only with action blocked." } }, "required": [ @@ -947,7 +947,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -972,15 +972,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective for action edit; ignored for other actions." + "description": "Replacement objective; valid only with action edit." }, "max_goal_rounds": { "type": "number", - "description": "Replacement round cap for action edit; ignored for other actions." + "description": "Replacement cap; valid only with action edit." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." + "description": "Concrete blocking condition; required only with action blocked." } }, "required": [ diff --git a/examples/acp-agent/tests/snapshots/permission-switching/tool-schemas.expected.json b/examples/acp-agent/tests/snapshots/permission-switching/tool-schemas.expected.json index dac439b700..ef40784fa5 100644 --- a/examples/acp-agent/tests/snapshots/permission-switching/tool-schemas.expected.json +++ b/examples/acp-agent/tests/snapshots/permission-switching/tool-schemas.expected.json @@ -397,7 +397,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -422,15 +422,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective for action edit; ignored for other actions." + "description": "Replacement objective; valid only with action edit." }, "max_goal_rounds": { "type": "number", - "description": "Replacement round cap for action edit; ignored for other actions." + "description": "Replacement cap; valid only with action edit." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." + "description": "Concrete blocking condition; required only with action blocked." } }, "required": [ @@ -947,7 +947,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -972,15 +972,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective for action edit; ignored for other actions." + "description": "Replacement objective; valid only with action edit." }, "max_goal_rounds": { "type": "number", - "description": "Replacement round cap for action edit; ignored for other actions." + "description": "Replacement cap; valid only with action edit." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." + "description": "Concrete blocking condition; required only with action blocked." } }, "required": [ diff --git a/examples/acp-agent/tests/snapshots/plan-mode/tool-schemas.expected.json b/examples/acp-agent/tests/snapshots/plan-mode/tool-schemas.expected.json index dac439b700..ef40784fa5 100644 --- a/examples/acp-agent/tests/snapshots/plan-mode/tool-schemas.expected.json +++ b/examples/acp-agent/tests/snapshots/plan-mode/tool-schemas.expected.json @@ -397,7 +397,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -422,15 +422,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective for action edit; ignored for other actions." + "description": "Replacement objective; valid only with action edit." }, "max_goal_rounds": { "type": "number", - "description": "Replacement round cap for action edit; ignored for other actions." + "description": "Replacement cap; valid only with action edit." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." + "description": "Concrete blocking condition; required only with action blocked." } }, "required": [ @@ -947,7 +947,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -972,15 +972,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective for action edit; ignored for other actions." + "description": "Replacement objective; valid only with action edit." }, "max_goal_rounds": { "type": "number", - "description": "Replacement round cap for action edit; ignored for other actions." + "description": "Replacement cap; valid only with action edit." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." + "description": "Concrete blocking condition; required only with action blocked." } }, "required": [ diff --git a/examples/acp-agent/tests/snapshots/pty-tools/tool-schemas.expected.json b/examples/acp-agent/tests/snapshots/pty-tools/tool-schemas.expected.json index 0ba75afbdc..529b1419da 100644 --- a/examples/acp-agent/tests/snapshots/pty-tools/tool-schemas.expected.json +++ b/examples/acp-agent/tests/snapshots/pty-tools/tool-schemas.expected.json @@ -526,7 +526,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -551,15 +551,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective for action edit; ignored for other actions." + "description": "Replacement objective; valid only with action edit." }, "max_goal_rounds": { "type": "number", - "description": "Replacement round cap for action edit; ignored for other actions." + "description": "Replacement cap; valid only with action edit." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." + "description": "Concrete blocking condition; required only with action blocked." } }, "required": [ diff --git a/examples/acp-agent/tests/snapshots/skill-load/tool-schemas.expected.json b/examples/acp-agent/tests/snapshots/skill-load/tool-schemas.expected.json index f3739bbc42..b01e7683d1 100644 --- a/examples/acp-agent/tests/snapshots/skill-load/tool-schemas.expected.json +++ b/examples/acp-agent/tests/snapshots/skill-load/tool-schemas.expected.json @@ -397,7 +397,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -422,15 +422,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective for action edit; ignored for other actions." + "description": "Replacement objective; valid only with action edit." }, "max_goal_rounds": { "type": "number", - "description": "Replacement round cap for action edit; ignored for other actions." + "description": "Replacement cap; valid only with action edit." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." + "description": "Concrete blocking condition; required only with action blocked." } }, "required": [ diff --git a/examples/acp-agent/tests/snapshots/text-turn/tool-schemas.expected.json b/examples/acp-agent/tests/snapshots/text-turn/tool-schemas.expected.json index f3739bbc42..b01e7683d1 100644 --- a/examples/acp-agent/tests/snapshots/text-turn/tool-schemas.expected.json +++ b/examples/acp-agent/tests/snapshots/text-turn/tool-schemas.expected.json @@ -397,7 +397,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -422,15 +422,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective for action edit; ignored for other actions." + "description": "Replacement objective; valid only with action edit." }, "max_goal_rounds": { "type": "number", - "description": "Replacement round cap for action edit; ignored for other actions." + "description": "Replacement cap; valid only with action edit." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." + "description": "Concrete blocking condition; required only with action blocked." } }, "required": [ diff --git a/examples/acp-agent/tests/snapshots/workspace-context/tool-schemas.expected.json b/examples/acp-agent/tests/snapshots/workspace-context/tool-schemas.expected.json index f3739bbc42..b01e7683d1 100644 --- a/examples/acp-agent/tests/snapshots/workspace-context/tool-schemas.expected.json +++ b/examples/acp-agent/tests/snapshots/workspace-context/tool-schemas.expected.json @@ -397,7 +397,7 @@ }, { "name": "update_goal", - "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", + "description": "Update the exact current goal revision. edit, pause, and resume require a direct top-level human request. During an automatic continuation of the current goal, complete and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.", "parameters": { "type": "object", "properties": { @@ -422,15 +422,15 @@ }, "objective": { "type": "string", - "description": "Replacement objective for action edit; ignored for other actions." + "description": "Replacement objective; valid only with action edit." }, "max_goal_rounds": { "type": "number", - "description": "Replacement round cap for action edit; ignored for other actions." + "description": "Replacement cap; valid only with action edit." }, "blocked_reason": { "type": "string", - "description": "Concrete blocking condition for action blocked; required for blocked, ignored for other actions." + "description": "Concrete blocking condition; required only with action blocked." } }, "required": [ diff --git a/packages/goal/tool-goal/README.md b/packages/goal/tool-goal/README.md index 4fc7cb0ea9..3f286f8ec3 100644 --- a/packages/goal/tool-goal/README.md +++ b/packages/goal/tool-goal/README.md @@ -6,7 +6,7 @@ The model-facing control surface for [`ctx.goals`](../goal/README.md): `get_goal - `get_goal()` returns the current goal or `null`, including the compare-and-set id/revision, durable phase, admitted/capped goal rounds, any blocker reason, and current process-local activation. - `create_goal(objective, max_goal_rounds?)` creates one goal from a direct top-level human turn. The model may infer long-running goal intent without an exact command phrase; non-human turns and subagents are rejected at execution. -- `update_goal(goal_id, revision, action, objective?, max_goal_rounds?, blocked_reason?)` supports `edit`, `pause`, `resume`, `complete`, and `blocked`. `edit` consumes replacement fields, `blocked` requires and persists `blocked_reason` with the stable code `model-reported`, and every action ignores fields it does not consume so providers that populate every schema property cannot prevent a valid transition. +- `update_goal(goal_id, revision, action, objective?, max_goal_rounds?, blocked_reason?)` supports `edit`, `pause`, `resume`, `complete`, and `blocked`. Replacements belong only to `edit`; `blocked_reason` is required only for `blocked` and is persisted with the stable code `model-reported`. All calls are exclusive, so a model-ordered batch observes earlier mutations and their new revisions. ACP and other clients receive pure generic cards: read for `get_goal`, other for mutations. diff --git a/packages/goal/tool-goal/src/index.ts b/packages/goal/tool-goal/src/index.ts index 5be69a9978..009a00376f 100644 --- a/packages/goal/tool-goal/src/index.ts +++ b/packages/goal/tool-goal/src/index.ts @@ -248,9 +248,8 @@ export function apply(ctx: Context, config: Config): void { name: 'update_goal', description: 'Update the exact current goal revision. edit, pause, and resume require a direct ' + 'top-level human request. During an automatic continuation of the current goal, complete ' - + 'and blocked are also allowed. Fields not used by the selected action are ignored. blocked is rejected before the configured ' - + 'minimum round count; the model remains responsible for judging that the same condition persisted across those rounds and ' - + 'must explain it in blocked_reason.', + + 'and blocked are also allowed. blocked is rejected before the configured minimum round count; the model remains ' + + 'responsible for judging that the same condition persisted across those rounds and must explain it in blocked_reason.', parameters: { goal_id: { type: 'string', required: true, description: 'Exact id returned by get_goal.' }, revision: { type: 'number', required: true, description: 'Exact positive revision returned by get_goal.' }, @@ -260,11 +259,11 @@ export function apply(ctx: Context, config: Config): void { enum: UPDATE_ACTIONS, description: 'edit | pause | resume | complete | blocked', }, - objective: { type: 'string', description: 'Replacement objective for action edit; ignored for other actions.' }, - max_goal_rounds: { type: 'number', description: 'Replacement round cap for action edit; ignored for other actions.' }, + objective: { type: 'string', description: 'Replacement objective; valid only with action edit.' }, + max_goal_rounds: { type: 'number', description: 'Replacement cap; valid only with action edit.' }, blocked_reason: { type: 'string', - description: 'Concrete blocking condition for action blocked; required for blocked, ignored for other actions.', + description: 'Concrete blocking condition; required only with action blocked.', }, }, output: GOAL_OUTPUT, @@ -277,12 +276,21 @@ export function apply(ctx: Context, config: Config): void { } if (args.action === 'edit') { requireDirectHuman(ctx, execution) + if (args.blocked_reason !== undefined) { + throw new HarnessError('blocked_reason is valid only with action blocked', 'GOAL_TOOL_INVALID_UPDATE') + } const goal = ctx.goals.edit(execution.agent, ref, replacements) observeMutation(terminalTurns, execution, false) return Promise.resolve(goalValue(goal)) } if (args.action === 'pause' || args.action === 'resume') { requireDirectHuman(ctx, execution) + if (args.objective !== undefined || args.max_goal_rounds !== undefined || args.blocked_reason !== undefined) { + throw new HarnessError( + 'objective and max_goal_rounds are valid only with action edit; blocked_reason is valid only with action blocked', + 'GOAL_TOOL_INVALID_UPDATE', + ) + } const goal = args.action === 'pause' ? ctx.goals.pause(execution.agent, ref) : ctx.goals.resume(execution.agent, ref) @@ -290,6 +298,15 @@ export function apply(ctx: Context, config: Config): void { return Promise.resolve(goalValue(goal)) } const authority = completionAuthority(ctx, execution) + if (args.objective !== undefined || args.max_goal_rounds !== undefined) { + throw new HarnessError( + 'objective and max_goal_rounds are valid only with action edit', + 'GOAL_TOOL_INVALID_UPDATE', + ) + } + if (args.action === 'complete' && args.blocked_reason !== undefined) { + throw new HarnessError('blocked_reason is valid only with action blocked', 'GOAL_TOOL_INVALID_UPDATE') + } if (args.action === 'blocked' && (args.blocked_reason === undefined || args.blocked_reason.trim().length === 0)) { throw new HarnessError('blocked_reason is required with action blocked', 'GOAL_TOOL_INVALID_UPDATE') diff --git a/packages/goal/tool-goal/tests/tool-goal.spec.ts b/packages/goal/tool-goal/tests/tool-goal.spec.ts index 30270fd34d..7b2ccdc347 100644 --- a/packages/goal/tool-goal/tests/tool-goal.spec.ts +++ b/packages/goal/tool-goal/tests/tool-goal.spec.ts @@ -377,12 +377,26 @@ describe('goal tool state transitions', () => { closeTurn(root, turn) }) - it('returns structured domain and required-argument failures', async () => { + it('returns structured domain and conditional-argument failures', async () => { const { ctx, root } = await harness() openTurn(root, { kind: 'user' }) const invalidCreate = await execute(ctx, 'create_goal', { objective: ' ' }, root.agent) expect(invalidCreate.error?.info?.code).toBe('GOAL_INVALID_OBJECTIVE') const created = ctx.goals.create(root.agent, { objective: 'valid' }) + const replacement = await execute(ctx, 'update_goal', { + goal_id: created.id, + revision: created.revision, + action: 'pause', + objective: 'not valid for pause', + }, root.agent) + expect(replacement.error?.info?.code).toBe('GOAL_TOOL_INVALID_UPDATE') + const terminalUpdate = await execute(ctx, 'update_goal', { + goal_id: created.id, + revision: created.revision, + action: 'complete', + max_goal_rounds: 2, + }, root.agent) + expect(terminalUpdate.error?.info?.code).toBe('GOAL_TOOL_INVALID_UPDATE') const blockedWithoutReason = await execute(ctx, 'update_goal', { goal_id: created.id, revision: created.revision, action: 'blocked', }, root.agent) @@ -391,45 +405,24 @@ describe('goal tool state transitions', () => { goal_id: created.id, revision: created.revision, action: 'blocked', blocked_reason: ' ', }, root.agent) expect(blockedWithEmptyReason.error?.info?.code).toBe('GOAL_TOOL_INVALID_UPDATE') + const completeWithReason = await execute(ctx, 'update_goal', { + goal_id: created.id, revision: created.revision, action: 'complete', blocked_reason: 'Not a blocker.', + }, root.agent) + expect(completeWithReason.error?.info?.code).toBe('GOAL_TOOL_INVALID_UPDATE') + const editWithReason = await execute(ctx, 'update_goal', { + goal_id: created.id, + revision: created.revision, + action: 'edit', + objective: 'still valid', + blocked_reason: 'Not valid for edit.', + }, root.agent) + expect(editWithReason.error?.info?.code).toBe('GOAL_TOOL_INVALID_UPDATE') const malformedRef = await execute(ctx, 'update_goal', { goal_id: '', revision: 0, action: 'edit', objective: 'x', - max_goal_rounds: 0, blocked_reason: '', }, root.agent) expect(malformedRef.error?.info?.code).toBe('GOAL_TOOL_INVALID_UPDATE') }) - it('ignores unused action fields emitted by strict-schema providers', async () => { - const { ctx, root } = await harness() - openTurn(root, { kind: 'user' }) - let goal = ctx.goals.create(root.agent, { objective: 'valid' }) - const strictArgs = { - objective: '', - max_goal_rounds: 0, - blocked_reason: '', - } - - const paused = await execute(ctx, 'update_goal', { - goal_id: goal.id, revision: goal.revision, action: 'pause', ...strictArgs, - }, root.agent) - goal = ctx.goals.get(root.agent)! - expect(resultGoal(paused)).toMatchObject({ phase: 'paused', objective: 'valid' }) - const resumed = await execute(ctx, 'update_goal', { - goal_id: goal.id, revision: goal.revision, action: 'resume', ...strictArgs, - }, root.agent) - goal = ctx.goals.get(root.agent)! - expect(resultGoal(resumed)).toMatchObject({ phase: 'active', objective: 'valid' }) - const edited = await execute(ctx, 'update_goal', { - goal_id: goal.id, revision: goal.revision, action: 'edit', objective: 'edited', - max_goal_rounds: 7, blocked_reason: '', - }, root.agent) - goal = ctx.goals.get(root.agent)! - expect(resultGoal(edited)).toMatchObject({ objective: 'edited', maxGoalRounds: 7 }) - const complete = await execute(ctx, 'update_goal', { - goal_id: goal.id, revision: goal.revision, action: 'complete', ...strictArgs, - }, root.agent) - expect(resultGoal(complete)).toMatchObject({ phase: 'complete', objective: 'edited', maxGoalRounds: 7 }) - }) - it('allows exact goal rounds to complete but not edit or pause', async () => { const { ctx, root } = await harness() const humanTurn = openTurn(root, { kind: 'user' }) From d436074cc23f2b304ae97198b968f25e7fb14c86 Mon Sep 17 00:00:00 2001 From: Turtle Date: Fri, 24 Jul 2026 13:15:39 +0800 Subject: [PATCH 5/8] fix(goal): accept empty update fillers --- packages/goal/tool-goal/src/index.ts | 35 ++++++++++- .../goal/tool-goal/tests/tool-goal.spec.ts | 59 +++++++++++++++++++ 2 files changed, 92 insertions(+), 2 deletions(-) diff --git a/packages/goal/tool-goal/src/index.ts b/packages/goal/tool-goal/src/index.ts index 009a00376f..91ca254b2d 100644 --- a/packages/goal/tool-goal/src/index.ts +++ b/packages/goal/tool-goal/src/index.ts @@ -132,6 +132,32 @@ function resolveConfig(config: Config): ResolvedConfig { return { blockedAfterConsecutiveRounds: blockedAfter } } +/** Remove only empty provider fillers from fields unused by the selected action. */ +function normalizeUpdateArgs(args: Record): Record { + const normalized = { ...args } + const empty = (value: unknown): boolean => value === undefined || value === null || value === '' || value === 0 + const removeEmpty = (key: string): void => { + if (empty(normalized[key])) normalized[key] = undefined + } + switch (normalized['action']) { + case 'edit': + removeEmpty('blocked_reason') + break + case 'blocked': + removeEmpty('objective') + removeEmpty('max_goal_rounds') + break + case 'pause': + case 'resume': + case 'complete': + removeEmpty('objective') + removeEmpty('max_goal_rounds') + removeEmpty('blocked_reason') + break + } + return normalized +} + /** Build the exact compare-and-set ref from model arguments. */ function goalRef(goalId: string, revision: number): GoalRef { if (goalId.length === 0 || goalId !== goalId.trim() @@ -244,7 +270,7 @@ export function apply(ctx: Context, config: Config): void { presentCall: args => present('Create goal', 'other', args.objective), })) - ctx.tools.register(defineTool({ + const updateGoal = defineTool({ name: 'update_goal', description: 'Update the exact current goal revision. edit, pause, and resume require a direct ' + 'top-level human request. During an automatic continuation of the current goal, complete ' @@ -333,5 +359,10 @@ export function apply(ctx: Context, config: Config): void { 'other', args.blocked_reason ?? args.objective ?? args.goal_id, ), - })) + }) + // Object-literal execute methods do not use `this`; retaining the reference is safe. + // eslint-disable-next-line @typescript-eslint/unbound-method + const executeUpdate = updateGoal.execute + updateGoal.execute = (args, exec) => executeUpdate(normalizeUpdateArgs(args as Record), exec) + ctx.tools.register(updateGoal) } diff --git a/packages/goal/tool-goal/tests/tool-goal.spec.ts b/packages/goal/tool-goal/tests/tool-goal.spec.ts index 7b2ccdc347..a1261d3cb8 100644 --- a/packages/goal/tool-goal/tests/tool-goal.spec.ts +++ b/packages/goal/tool-goal/tests/tool-goal.spec.ts @@ -423,6 +423,65 @@ describe('goal tool state transitions', () => { expect(malformedRef.error?.info?.code).toBe('GOAL_TOOL_INVALID_UPDATE') }) + it('accepts only empty fillers in fields unused by the selected action', async () => { + const { ctx, root } = await harness() + openTurn(root, { kind: 'user' }) + let goal = ctx.goals.create(root.agent, { objective: 'valid' }) + + const edited = await execute(ctx, 'update_goal', { + goal_id: goal.id, + revision: goal.revision, + action: 'edit', + objective: 'edited', + blocked_reason: '', + }, root.agent) + expect(resultGoal(edited)).toMatchObject({ objective: 'edited' }) + goal = ctx.goals.get(root.agent)! + + const paused = await execute(ctx, 'update_goal', { + goal_id: goal.id, + revision: goal.revision, + action: 'pause', + objective: '', + max_goal_rounds: 0, + blocked_reason: null, + }, root.agent) + expect(resultGoal(paused)).toMatchObject({ phase: 'paused', objective: 'edited' }) + goal = ctx.goals.get(root.agent)! + + const resumed = await execute(ctx, 'update_goal', { + goal_id: goal.id, + revision: goal.revision, + action: 'resume', + objective: null, + max_goal_rounds: '', + blocked_reason: '', + }, root.agent) + expect(resultGoal(resumed)).toMatchObject({ phase: 'active', objective: 'edited' }) + goal = ctx.goals.get(root.agent)! + + const blocked = await execute(ctx, 'update_goal', { + goal_id: goal.id, + revision: goal.revision, + action: 'blocked', + objective: '', + max_goal_rounds: null, + blocked_reason: 'actual blocker', + }, root.agent) + expect(resultGoal(blocked)).toMatchObject({ phase: 'blocked' }) + goal = ctx.goals.resume(root.agent, { id: goal.id, revision: goal.revision + 1 }) + + const complete = await execute(ctx, 'update_goal', { + goal_id: goal.id, + revision: goal.revision, + action: 'complete', + objective: '', + max_goal_rounds: 0, + blocked_reason: '', + }, root.agent) + expect(resultGoal(complete)).toMatchObject({ phase: 'complete', objective: 'edited' }) + }) + it('allows exact goal rounds to complete but not edit or pause', async () => { const { ctx, root } = await harness() const humanTurn = openTurn(root, { kind: 'user' }) From d94a4916f1c8a43a2c4b3c194bd183814edf52d0 Mon Sep 17 00:00:00 2001 From: Turtle Date: Fri, 24 Jul 2026 13:16:14 +0800 Subject: [PATCH 6/8] fix(goal): preserve normalized argument shape --- packages/goal/tool-goal/src/index.ts | 26 +++++--------------------- 1 file changed, 5 insertions(+), 21 deletions(-) diff --git a/packages/goal/tool-goal/src/index.ts b/packages/goal/tool-goal/src/index.ts index 91ca254b2d..a0df65a6ed 100644 --- a/packages/goal/tool-goal/src/index.ts +++ b/packages/goal/tool-goal/src/index.ts @@ -134,28 +134,12 @@ function resolveConfig(config: Config): ResolvedConfig { /** Remove only empty provider fillers from fields unused by the selected action. */ function normalizeUpdateArgs(args: Record): Record { - const normalized = { ...args } + const action = args['action'] const empty = (value: unknown): boolean => value === undefined || value === null || value === '' || value === 0 - const removeEmpty = (key: string): void => { - if (empty(normalized[key])) normalized[key] = undefined - } - switch (normalized['action']) { - case 'edit': - removeEmpty('blocked_reason') - break - case 'blocked': - removeEmpty('objective') - removeEmpty('max_goal_rounds') - break - case 'pause': - case 'resume': - case 'complete': - removeEmpty('objective') - removeEmpty('max_goal_rounds') - removeEmpty('blocked_reason') - break - } - return normalized + const unused = (key: string): boolean => key === 'blocked_reason' + ? action !== 'blocked' + : (key === 'objective' || key === 'max_goal_rounds') && action !== 'edit' + return Object.fromEntries(Object.entries(args).filter(([key, value]) => !unused(key) || !empty(value))) } /** Build the exact compare-and-set ref from model arguments. */ From d3e24356ba5ade49bb44bc4b8dcbc929c8cb0a73 Mon Sep 17 00:00:00 2001 From: Turtle Date: Fri, 24 Jul 2026 13:53:06 +0800 Subject: [PATCH 7/8] refactor(goal): loosen executor presence checks --- packages/goal/tool-goal/src/index.ts | 35 +++++++++---------- .../goal/tool-goal/tests/tool-goal.spec.ts | 8 ++--- 2 files changed, 20 insertions(+), 23 deletions(-) diff --git a/packages/goal/tool-goal/src/index.ts b/packages/goal/tool-goal/src/index.ts index a0df65a6ed..96e6389f85 100644 --- a/packages/goal/tool-goal/src/index.ts +++ b/packages/goal/tool-goal/src/index.ts @@ -132,14 +132,14 @@ function resolveConfig(config: Config): ResolvedConfig { return { blockedAfterConsecutiveRounds: blockedAfter } } -/** Remove only empty provider fillers from fields unused by the selected action. */ -function normalizeUpdateArgs(args: Record): Record { - const action = args['action'] - const empty = (value: unknown): boolean => value === undefined || value === null || value === '' || value === 0 - const unused = (key: string): boolean => key === 'blocked_reason' - ? action !== 'blocked' - : (key === 'objective' || key === 'max_goal_rounds') && action !== 'edit' - return Object.fromEntries(Object.entries(args).filter(([key, value]) => !unused(key) || !empty(value))) +/** Whether an optional string carries a meaningful action-specific value. */ +function hasText(value: string | undefined): boolean { + return value !== undefined && value !== '' +} + +/** Whether an optional round cap carries a meaningful action-specific value. */ +function hasRoundCap(value: number | undefined): boolean { + return value !== undefined && value !== 0 } /** Build the exact compare-and-set ref from model arguments. */ @@ -254,7 +254,7 @@ export function apply(ctx: Context, config: Config): void { presentCall: args => present('Create goal', 'other', args.objective), })) - const updateGoal = defineTool({ + ctx.tools.register(defineTool({ name: 'update_goal', description: 'Update the exact current goal revision. edit, pause, and resume require a direct ' + 'top-level human request. During an automatic continuation of the current goal, complete ' @@ -286,7 +286,9 @@ export function apply(ctx: Context, config: Config): void { } if (args.action === 'edit') { requireDirectHuman(ctx, execution) - if (args.blocked_reason !== undefined) { + // Some strict-schema providers emit empty placeholders for every declared + // optional field. Reject only values that could change another action. + if (hasText(args.blocked_reason)) { throw new HarnessError('blocked_reason is valid only with action blocked', 'GOAL_TOOL_INVALID_UPDATE') } const goal = ctx.goals.edit(execution.agent, ref, replacements) @@ -295,7 +297,7 @@ export function apply(ctx: Context, config: Config): void { } if (args.action === 'pause' || args.action === 'resume') { requireDirectHuman(ctx, execution) - if (args.objective !== undefined || args.max_goal_rounds !== undefined || args.blocked_reason !== undefined) { + if (hasText(args.objective) || hasRoundCap(args.max_goal_rounds) || hasText(args.blocked_reason)) { throw new HarnessError( 'objective and max_goal_rounds are valid only with action edit; blocked_reason is valid only with action blocked', 'GOAL_TOOL_INVALID_UPDATE', @@ -308,13 +310,13 @@ export function apply(ctx: Context, config: Config): void { return Promise.resolve(goalValue(goal)) } const authority = completionAuthority(ctx, execution) - if (args.objective !== undefined || args.max_goal_rounds !== undefined) { + if (hasText(args.objective) || hasRoundCap(args.max_goal_rounds)) { throw new HarnessError( 'objective and max_goal_rounds are valid only with action edit', 'GOAL_TOOL_INVALID_UPDATE', ) } - if (args.action === 'complete' && args.blocked_reason !== undefined) { + if (args.action === 'complete' && hasText(args.blocked_reason)) { throw new HarnessError('blocked_reason is valid only with action blocked', 'GOAL_TOOL_INVALID_UPDATE') } if (args.action === 'blocked' @@ -343,10 +345,5 @@ export function apply(ctx: Context, config: Config): void { 'other', args.blocked_reason ?? args.objective ?? args.goal_id, ), - }) - // Object-literal execute methods do not use `this`; retaining the reference is safe. - // eslint-disable-next-line @typescript-eslint/unbound-method - const executeUpdate = updateGoal.execute - updateGoal.execute = (args, exec) => executeUpdate(normalizeUpdateArgs(args as Record), exec) - ctx.tools.register(updateGoal) + })) } diff --git a/packages/goal/tool-goal/tests/tool-goal.spec.ts b/packages/goal/tool-goal/tests/tool-goal.spec.ts index a1261d3cb8..cf48b3dcbc 100644 --- a/packages/goal/tool-goal/tests/tool-goal.spec.ts +++ b/packages/goal/tool-goal/tests/tool-goal.spec.ts @@ -444,7 +444,7 @@ describe('goal tool state transitions', () => { action: 'pause', objective: '', max_goal_rounds: 0, - blocked_reason: null, + blocked_reason: '', }, root.agent) expect(resultGoal(paused)).toMatchObject({ phase: 'paused', objective: 'edited' }) goal = ctx.goals.get(root.agent)! @@ -453,8 +453,8 @@ describe('goal tool state transitions', () => { goal_id: goal.id, revision: goal.revision, action: 'resume', - objective: null, - max_goal_rounds: '', + objective: '', + max_goal_rounds: 0, blocked_reason: '', }, root.agent) expect(resultGoal(resumed)).toMatchObject({ phase: 'active', objective: 'edited' }) @@ -465,7 +465,7 @@ describe('goal tool state transitions', () => { revision: goal.revision, action: 'blocked', objective: '', - max_goal_rounds: null, + max_goal_rounds: 0, blocked_reason: 'actual blocker', }, root.agent) expect(resultGoal(blocked)).toMatchObject({ phase: 'blocked' }) From 9e691d54c09173f8074f134008e1eeda7920b491 Mon Sep 17 00:00:00 2001 From: Tianyi Cui <53024+tianyicui@users.noreply.github.com> Date: Fri, 24 Jul 2026 20:45:23 +0800 Subject: [PATCH 8/8] fix(goal): normalize strict-schema fillers --- ...26-07-19-model-facing-goal-tools.i18n.yaml | 4 +- .../2026-07-19-model-facing-goal-tools.md | 8 ++- .../2026-07-19-model-facing-goal-tools.zh.md | 8 ++- .../headless-agent/tests/headless.snapshot.ts | 7 +- .../tests/snapshots/goal-tools/input.json | 2 +- .../snapshots/goal-tools/replay.override.json | 10 +++ .../goal-tools/stream-json.expected.jsonl | 66 +++++++++++-------- packages/goal/tool-goal/README.md | 4 +- packages/goal/tool-goal/src/index.ts | 20 +++--- .../goal/tool-goal/tests/tool-goal.spec.ts | 21 ++++++ 10 files changed, 101 insertions(+), 49 deletions(-) diff --git a/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.i18n.yaml b/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.i18n.yaml index e53c591aa5..70ca07b414 100644 --- a/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.i18n.yaml +++ b/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.i18n.yaml @@ -2,5 +2,5 @@ # side as of the last confirmed-consistent state. Both languages carry equal authority; # after editing either side, bring the other along and re-record with: # pnpm run verify-translation-pairing --write -2026-07-19-model-facing-goal-tools.md: 7cc3907d708115207e166455ea988120a03d768b -2026-07-19-model-facing-goal-tools.zh.md: 1a381160354d6a2a24f957f41bc9e375c1ab01ca +2026-07-19-model-facing-goal-tools.md: 286329390a058c0302520fd2203e5becb8c81395 +2026-07-19-model-facing-goal-tools.zh.md: b0b4fc99ada3597fbab58081f52309e21dd43bac diff --git a/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.md b/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.md index 7cc3907d70..286329390a 100644 --- a/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.md +++ b/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.md @@ -16,11 +16,11 @@ The surface also needs to preserve the separation between durable state and live ### Tools and model contract -`get_goal()` returns the current goal or `null`. A non-null result contains the compare-and-set id and revision, objective, durable phase, admitted and maximum goal rounds, any blocker reason, plus the process-local activation observation. `create_goal(objective, max_goal_rounds?)` creates one long-running same-session objective. `update_goal(goal_id, revision, action, objective?, max_goal_rounds?, blocked_reason?)` supports `edit`, `pause`, `resume`, `complete`, and `blocked`; replacement fields are valid only for `edit`, while a non-empty `blocked_reason` is required only for `blocked` and persists under the stable `model-reported` code. +`get_goal()` returns the current goal or `null`. A non-null result contains the compare-and-set id and revision, objective, durable phase, admitted and maximum goal rounds, any blocker reason, plus the process-local activation observation. `create_goal(objective, max_goal_rounds?)` creates one long-running same-session objective. `update_goal(goal_id, revision, action, objective?, max_goal_rounds?, blocked_reason?)` supports `edit`, `pause`, `resume`, `complete`, and `blocked`; replacement fields are valid only for `edit`, while a non-empty `blocked_reason` is required only for `blocked` and persists under the stable `model-reported` code. The executor treats exact empty-string optional fields and a zero `max_goal_rounds` as strict-schema fillers: they count as omitted, an edit still requires at least one meaningful replacement, and all non-filler values retain the action restrictions. The prompt tells the model that it may infer goal intent from a direct human request in any wording or language, but should not convert routine single-turn work into a goal. It must read the current goal before updating and copy the exact id and revision. On a restored or forked active-but-disarmed goal, a semantic human request to continue is grounds for `resume`. Completion is reserved for an achieved objective, and difficulty or uncertainty alone is not a blocker; a block report must name the concrete condition. -All three tools use exclusive execution so a model-ordered batch observes prior mutations and their new revisions. Results are compact JSON. ACP presentation is a pure function of arguments and uses generic read or mutation cards; activation is reported only as live observation and is never written into replay state. +All three tools use exclusive execution so a model-ordered batch observes prior mutations and their new revisions. Results are compact JSON. ACP presentation is a pure function of arguments and uses generic read or mutation cards; mutation cards select meaningful action values before the goal id, so accepted fillers cannot blank their input. Activation is reported only as live observation and is never written into replay state. An autonomous goal round that successfully reports completion or blocking contributes the existing terminal `agent/turn-stop` decision for that physical turn, preventing an unnecessary follow-up request. Direct-human mutations do not contribute a terminal stop: the assistant can acknowledge the change, and concurrent human steering remains available to ordinary continuation folding. @@ -38,7 +38,7 @@ Complete and blocked accept either direct-human authority or the exact current g ## Testing -Unit coverage pins registration and disposal, exclusive scheduling, generated prompt policy, generic presentation, direct-human creation in a non-English turn, exact/stale/non-running agent and driver checks, live-child rejection, resumed-fork root authority, steering, mismatched initiators, read/create/edit/pause/resume behavior, conditional blocker explanations, rearming after a session-start edge, authority-before-conditional-argument failures, exact goal-round completion, autonomous-only terminal stopping, the configured blocking threshold, and immediate human blocking. A keyless replay snapshot mounts the goal domain and tools into the real headless one-shot application, drives `create_goal` and `get_goal` through the shipped loop and persistence stack, pins its stream-json transcript, and inspects the externally persisted goal change. The echo-agent fixture is intentionally not used as an application-UX surrogate. +Unit coverage pins registration and disposal, exclusive scheduling, generated prompt policy, filler-safe generic presentation, direct-human creation in a non-English turn, exact/stale/non-running agent and driver checks, live-child rejection, resumed-fork root authority, steering, mismatched initiators, read/create/partial-edit/pause/resume behavior including strict-schema fillers, conditional blocker explanations, rearming after a session-start edge, authority-before-conditional-argument failures, exact goal-round completion, autonomous-only terminal stopping, the configured blocking threshold, and immediate human blocking. A keyless replay snapshot mounts the goal domain and tools into the real headless one-shot application, drives a strict-filler `update_goal` probe plus `create_goal` and `get_goal` through the shipped loop and persistence stack, pins its stream-json transcript, and inspects the externally persisted goal change. The echo-agent fixture is intentionally not used as an application-UX surrogate. ## Alternatives considered @@ -48,6 +48,7 @@ Unit coverage pins registration and disposal, exclusive scheduling, generated pr - **Authorize from persisted root or fork metadata** — rejected because a fork that becomes an independently resumed top-level session should accept new human authority, while a currently owned child should not. - **Let autonomous rounds edit or resume the goal** — rejected because continuation authority is narrower than authority to redefine or restart the human objective. - **Treat the blocked threshold as an evaluator** — rejected because event counts cannot prove that an obstacle is semantically unchanged or truly terminal. +- **Reject every present action-specific field** — rejected because strict-schema providers can serialize zero-value placeholders for every optional field; only meaningful values can express a conflicting action. ## Consequences @@ -56,6 +57,7 @@ Unit coverage pins registration and disposal, exclusive scheduling, generated pr - Human requests can create and rearm goals through ordinary natural language, while restored sessions remain inert until such input arrives. - Goal rounds can finish or report a repeated blocker but cannot broaden their own mandate. - Deployment policy selects the blocking lower bound; the same resolved value controls enforcement and prompt guidance. +- Strict-schema provider fillers interoperate without allowing meaningful cross-action updates. ## Known limitations and deferred work diff --git a/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.zh.md b/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.zh.md index 1a38116035..b0b4fc99ad 100644 --- a/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.zh.md +++ b/.agents/notes/implemented/feature/2026-07-19-model-facing-goal-tools.zh.md @@ -16,11 +16,11 @@ Status: implemented ### 工具与模型契约 -`get_goal()` 返回当前目标或 `null`。非空结果包含用于比较并交换的 id 与修订号、目标描述、持久阶段、已接纳和最大目标回合数、可能存在的阻塞原因，以及进程本地激活态观察。`create_goal(objective, max_goal_rounds?)` 创建一个长时间运行的同会话目标。`update_goal(goal_id, revision, action, objective?, max_goal_rounds?, blocked_reason?)` 支持 `edit`、`pause`、`resume`、`complete` 和 `blocked`；替换字段仅对 `edit` 有效，非空的 `blocked_reason` 仅在 `blocked` 时必填，并以稳定代码 `model-reported` 持久化。 +`get_goal()` 返回当前目标或 `null`。非空结果包含用于比较并交换的 id 与修订号、目标描述、持久阶段、已接纳和最大目标回合数、可能存在的阻塞原因，以及进程本地激活态观察。`create_goal(objective, max_goal_rounds?)` 创建一个长时间运行的同会话目标。`update_goal(goal_id, revision, action, objective?, max_goal_rounds?, blocked_reason?)` 支持 `edit`、`pause`、`resume`、`complete` 和 `blocked`；替换字段仅对 `edit` 有效，非空的 `blocked_reason` 仅在 `blocked` 时必填，并以稳定代码 `model-reported` 持久化。执行器把值恰好为空字符串的可选字段和值为 0 的 `max_goal_rounds` 视为严格 schema 占位值：这些值等同于省略；编辑时仍必须提供至少一个有实际意义的替换字段；所有非占位值仍受对应操作的限制。 提示词告诉模型：它可以从任何措辞或语言的直接人类请求中推断目标意图，但不应把常规单轮工作转换为目标。更新前必须读取当前目标，并复制准确的 id 和修订号。对于恢复或派生后处于活跃但未激活状态的目标，人类在语义上要求继续即可成为执行 `resume` 的依据。只有目标已经实现时才能标记完成，困难或不确定性本身不构成阻塞；阻塞报告必须说明具体条件。 -三个工具都采用独占执行，使模型排序的批次可以观察此前变更及其新修订号。结果为紧凑 JSON。ACP 展示是参数的纯函数，使用通用读取或变更卡片；激活态仅作为实时观察返回，绝不会写入回放状态。 +三个工具都采用独占执行，使模型排序的批次可以观察此前变更及其新修订号。结果为紧凑 JSON。ACP 展示是参数的纯函数，使用通用读取或变更卡片；变更卡片选择输入时，先取有实际意义的操作值，再取目标 id，因此允许的占位值不会使卡片输入留空。激活态仅作为实时观察返回，绝不会写入回放状态。 自主目标回合成功报告完成或阻塞后，插件会为该物理轮次贡献现有的终止型 `agent/turn-stop` 决策，避免再发起一次不必要的模型请求。直接人类发起的变更不会贡献终止决策：智能体可以确认该变更，并且并发的人类 steering（转向）仍可参与普通的继续执行折叠。 @@ -38,7 +38,7 @@ Status: implemented ## 测试 -单元测试固定注册与释放、独占调度、生成的提示词策略、通用展示、非英语轮次中的直接人类创建、精确/陈旧/非运行中智能体与驱动检查、实时子智能体拒绝、恢复后派生根的权限、steering、发起者不匹配、读取/创建/编辑/暂停/恢复行为、条件式阻塞说明、会话启动边沿后的重新激活、权限先于条件参数失败、准确目标回合的完成、仅自主回合触发终止、可配置阻塞阈值，以及人类立即阻塞。无密钥回放快照把目标领域和工具挂载到真实的 headless 单次运行应用中，通过随附循环与持久化栈驱动 `create_goal` 和 `get_goal`，固定 stream-json 转录，并检查外部持久化的目标变更。这里有意不把 echo-agent 测试夹具当作应用 UX 的替代品。 +单元测试固定注册与释放、独占调度、生成的提示词策略、可安全处理占位值的通用展示、非英语轮次中的直接人类创建、精确/陈旧/非运行中智能体与驱动检查、实时子智能体拒绝、恢复后派生根的权限、steering、发起者不匹配、读取/创建/部分字段编辑/暂停/恢复行为（包括严格 schema 占位值）、条件式阻塞说明、会话启动边沿后的重新激活、权限先于条件参数失败、准确目标回合的完成、仅自主回合触发终止、可配置阻塞阈值，以及人类立即阻塞。无密钥回放快照把目标领域和工具挂载到真实的 headless 单次运行应用中，通过随附循环与持久化栈驱动一次携带严格 schema 占位值的 `update_goal` 探测，以及对 `create_goal` 和 `get_goal` 的调用，固定 stream-json 转录，并检查外部持久化的目标变更。这里有意不把 echo-agent 测试夹具当作应用 UX 的替代品。 ## 考虑过的替代方案 @@ -48,6 +48,7 @@ Status: implemented - **根据持久的根或派生元数据授权**——不予采纳，因为成为独立恢复顶层会话的派生应接受新的人类权限，而当前仍受所有权约束的子智能体则不应接受。 - **允许自主回合编辑或恢复目标**——不予采纳，因为继续执行权限比重新定义或重启人类目标的权限更窄。 - **把阻塞阈值当作评估器**——不予采纳，因为事件计数无法证明障碍在语义上未改变或确实不可继续。 +- **拒绝所有已提供的特定操作字段**——不予采纳，因为采用严格 schema 的提供方可能为每个可选字段序列化零值占位符；只有有实际意义的字段值才能表示与指定操作相冲突的另一项操作。 ## 后果 @@ -56,6 +57,7 @@ Status: implemented - 人类可以通过普通自然语言请求创建和重新激活目标，而恢复后的会话在收到此类输入前保持静止。 - 目标回合可以完成或报告重复阻塞，但不能自行扩大任务权限。 - 部署策略选择阻塞下限；同一个解析后的值同时控制执行与提示词指导。 +- 系统可兼容采用严格 schema 的提供方所填入的占位值，同时不会放行有实际意义的跨操作更新。 ## 已知限制与延期工作 diff --git a/examples/headless-agent/tests/headless.snapshot.ts b/examples/headless-agent/tests/headless.snapshot.ts index a1ce4b6cc5..6852fed20b 100644 --- a/examples/headless-agent/tests/headless.snapshot.ts +++ b/examples/headless-agent/tests/headless.snapshot.ts @@ -222,7 +222,12 @@ describe('headless stream-json snapshots', () => { const records = parseJsonl(logs[0]?.content ?? '') const calls = records.filter(record => record.type === 'tool/call') .map(record => (record.data as JsonObject | undefined)?.name) - expect(calls).toEqual(['create_goal', 'get_goal']) + expect(calls).toEqual(['update_goal', 'create_goal', 'get_goal']) + const probeResult = records.find(record => record.type === 'tool/result' + && (record.data as JsonObject | undefined)?.callId === 'call_goal_probe') + const probeData = probeResult?.data as JsonObject | undefined + expect(probeData?.isError).toBe(true) + expect((probeData?.error as JsonObject | undefined)?.code).toBe('GOAL_NOT_FOUND') const goalChanges = records.filter((record) => { if (record.type !== 'user/message') return false const data = record.data as JsonObject | undefined diff --git a/examples/headless-agent/tests/snapshots/goal-tools/input.json b/examples/headless-agent/tests/snapshots/goal-tools/input.json index 5263ccd4e2..8449d44c4c 100644 --- a/examples/headless-agent/tests/snapshots/goal-tools/input.json +++ b/examples/headless-agent/tests/snapshots/goal-tools/input.json @@ -2,7 +2,7 @@ "steps": [ { "op": "prompt", - "text": "Create a durable goal to finish the snapshot proof, then inspect it." + "text": "Probe strict-schema fillers against missing-goal revision 1, then create a durable goal to finish the snapshot proof and inspect it." } ] } diff --git a/examples/headless-agent/tests/snapshots/goal-tools/replay.override.json b/examples/headless-agent/tests/snapshots/goal-tools/replay.override.json index aec5204c7d..c4716ba7b0 100644 --- a/examples/headless-agent/tests/snapshots/goal-tools/replay.override.json +++ b/examples/headless-agent/tests/snapshots/goal-tools/replay.override.json @@ -1,4 +1,14 @@ [ + { + "kind": "chunks", + "chunks": [ + { "type": "block-start", "index": 0, "blockType": "tool-call" }, + { "type": "tool-call-delta", "index": 0, "id": "call_goal_probe", "name": "update_goal", "argumentsDelta": "{\"goal_id\":\"missing-goal\",\"revision\":1,\"action\":\"pause\",\"objective\":\"\",\"max_goal_rounds\":0,\"blocked_reason\":\"\"}" }, + { "type": "block-end", "index": 0, "block": { "type": "tool-call", "id": "call_goal_probe", "name": "update_goal", "arguments": "{\"goal_id\":\"missing-goal\",\"revision\":1,\"action\":\"pause\",\"objective\":\"\",\"max_goal_rounds\":0,\"blocked_reason\":\"\"}" } }, + { "type": "usage", "usage": { "inputTokens": 15, "outputTokens": 6 } }, + { "type": "finish", "reason": { "kind": "tool-calls" } } + ] + }, { "kind": "chunks", "chunks": [ diff --git a/examples/headless-agent/tests/snapshots/goal-tools/stream-json.expected.jsonl b/examples/headless-agent/tests/snapshots/goal-tools/stream-json.expected.jsonl index 5005192ec8..16518b2d03 100644 --- a/examples/headless-agent/tests/snapshots/goal-tools/stream-json.expected.jsonl +++ b/examples/headless-agent/tests/snapshots/goal-tools/stream-json.expected.jsonl @@ -1,35 +1,45 @@ {"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"turn/start","seq":0,"time":0,"data":{"turn":1,"trigger":{"kind":"message","source":{"kind":"user"}}}}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"user/message","seq":1,"time":0,"data":{"content":[{"type":"text","text":"Create a durable goal to finish the snapshot proof, then inspect it."}],"source":{"kind":"user"}},"surfaceOp":"append"}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"session/title","seq":2,"time":0,"data":{"title":"Create a durable goal to","messageSeqs":[1],"source":{"kind":"fallback"}}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"user/message","seq":1,"time":0,"data":{"content":[{"type":"text","text":"Probe strict-schema fillers against missing-goal revision 1, then create a durable goal to finish the snapshot proof and inspect it."}],"source":{"kind":"user"}},"surfaceOp":"append"}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"session/title","seq":2,"time":0,"data":{"title":"Probe strict-schema fillers against miss","messageSeqs":[1],"source":{"kind":"fallback"}}}} {"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"step/start","seq":3,"time":0,"data":{"turn":1,"step":1}}} {"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"request/header","seq":4,"time":0,"data":{"header":{"config":{"provider":"deepseek","model":"deepseek-v4-flash"},"system":"{{system}}","tools":"{{tools}}"},"reason":"initial"}}} {"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":5,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":6,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":0,"id":"call_goal_create","name":"create_goal","argumentsDelta":"{\"objective\":\"Finish the headless goal-tool snapshot proof\",\"max_goal_rounds\":7}"}}}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":7,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"call_goal_create","name":"create_goal","arguments":"{\"objective\":\"Finish the headless goal-tool snapshot proof\",\"max_goal_rounds\":7}"}}}}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":8,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":20,"outputTokens":8}}}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":6,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"tool-call-delta","index":0,"id":"call_goal_probe","name":"update_goal","argumentsDelta":"{\"goal_id\":\"missing-goal\",\"revision\":1,\"action\":\"pause\",\"objective\":\"\",\"max_goal_rounds\":0,\"blocked_reason\":\"\"}"}}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":7,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"call_goal_probe","name":"update_goal","arguments":"{\"goal_id\":\"missing-goal\",\"revision\":1,\"action\":\"pause\",\"objective\":\"\",\"max_goal_rounds\":0,\"blocked_reason\":\"\"}"}}}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":8,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":15,"outputTokens":6}}}}} {"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":9,"time":0,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/message","seq":10,"time":0,"data":{"turn":1,"step":1,"content":[{"type":"tool-call","id":"call_goal_create","name":"create_goal","arguments":"{\"objective\":\"Finish the headless goal-tool snapshot proof\",\"max_goal_rounds\":7}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":20,"outputTokens":8}},"sourceEventSeqs":[5,6,7,8,9],"surfaceOp":"append"}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"tool/call","seq":11,"time":0,"data":{"turn":1,"step":1,"callId":"call_goal_create","name":"create_goal","arguments":"{\"objective\":\"Finish the headless goal-tool snapshot proof\",\"max_goal_rounds\":7}"}}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"tool/result","seq":12,"time":0,"data":{"turn":1,"step":1,"callId":"call_goal_create","content":[{"type":"text","text":"{\"goal\":{\"id\":\"goal-{{sessionId}}\",\"revision\":1,\"objective\":\"Finish the headless goal-tool snapshot proof\",\"phase\":\"active\",\"roundsStarted\":0,\"maxGoalRounds\":7},\"activation\":\"armed\"}"}],"isError":false},"sourceEventSeqs":[11],"surfaceOp":"append"}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"user/message","seq":13,"time":0,"data":{"content":[{"type":"text","text":"{\"goal\":{\"id\":\"goal-{{sessionId}}\",\"revision\":1,\"objective\":\"Finish the headless goal-tool snapshot proof\",\"phase\":\"active\",\"maxGoalRounds\":7},\"roundsStarted\":0,\"createdAt\":0,\"updatedAt\":0}"}],"source":{"kind":"goal","goalId":"goal-{{sessionId}}","revision":1,"round":0},"meta":{"kind":"goal/change","version":1,"operation":"create","goal":{"id":"goal-{{sessionId}}","revision":1,"objective":"Finish the headless goal-tool snapshot proof","phase":"active","maxGoalRounds":7},"roundsStarted":0,"createdAt":0,"updatedAt":0}},"surfaceOp":"append"}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"step/end","seq":14,"time":0,"data":{"turn":1,"step":1}}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"step/start","seq":15,"time":0,"data":{"turn":1,"step":2}}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":16,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":17,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"tool-call-delta","index":0,"id":"call_goal_get","name":"get_goal","argumentsDelta":"{}"}}}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":18,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"call_goal_get","name":"get_goal","arguments":"{}"}}}}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":19,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":30,"outputTokens":4}}}}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":20,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/message","seq":21,"time":0,"data":{"turn":1,"step":2,"content":[{"type":"tool-call","id":"call_goal_get","name":"get_goal","arguments":"{}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":30,"outputTokens":4}},"sourceEventSeqs":[16,17,18,19,20],"surfaceOp":"append"}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"tool/call","seq":22,"time":0,"data":{"turn":1,"step":2,"callId":"call_goal_get","name":"get_goal","arguments":"{}"}}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"tool/result","seq":23,"time":0,"data":{"turn":1,"step":2,"callId":"call_goal_get","content":[{"type":"text","text":"{\"goal\":{\"id\":\"goal-{{sessionId}}\",\"revision\":1,\"objective\":\"Finish the headless goal-tool snapshot proof\",\"phase\":\"active\",\"roundsStarted\":0,\"maxGoalRounds\":7},\"activation\":\"armed\"}"}],"isError":false},"sourceEventSeqs":[22],"surfaceOp":"append"}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/message","seq":10,"time":0,"data":{"turn":1,"step":1,"content":[{"type":"tool-call","id":"call_goal_probe","name":"update_goal","arguments":"{\"goal_id\":\"missing-goal\",\"revision\":1,\"action\":\"pause\",\"objective\":\"\",\"max_goal_rounds\":0,\"blocked_reason\":\"\"}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":15,"outputTokens":6}},"sourceEventSeqs":[5,6,7,8,9],"surfaceOp":"append"}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"tool/call","seq":11,"time":0,"data":{"turn":1,"step":1,"callId":"call_goal_probe","name":"update_goal","arguments":"{\"goal_id\":\"missing-goal\",\"revision\":1,\"action\":\"pause\",\"objective\":\"\",\"max_goal_rounds\":0,\"blocked_reason\":\"\"}"}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"tool/result","seq":12,"time":0,"data":{"turn":1,"step":1,"callId":"call_goal_probe","content":[{"type":"text","text":"Error: no current goal"}],"isError":true,"error":{"name":"GoalError","code":"GOAL_NOT_FOUND"}},"sourceEventSeqs":[11],"surfaceOp":"append"}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"step/end","seq":13,"time":0,"data":{"turn":1,"step":1}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"step/start","seq":14,"time":0,"data":{"turn":1,"step":2}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":15,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":16,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"tool-call-delta","index":0,"id":"call_goal_create","name":"create_goal","argumentsDelta":"{\"objective\":\"Finish the headless goal-tool snapshot proof\",\"max_goal_rounds\":7}"}}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":17,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"call_goal_create","name":"create_goal","arguments":"{\"objective\":\"Finish the headless goal-tool snapshot proof\",\"max_goal_rounds\":7}"}}}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":18,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"usage","usage":{"inputTokens":20,"outputTokens":8}}}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":19,"time":0,"data":{"turn":1,"step":2,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/message","seq":20,"time":0,"data":{"turn":1,"step":2,"content":[{"type":"tool-call","id":"call_goal_create","name":"create_goal","arguments":"{\"objective\":\"Finish the headless goal-tool snapshot proof\",\"max_goal_rounds\":7}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":20,"outputTokens":8}},"sourceEventSeqs":[15,16,17,18,19],"surfaceOp":"append"}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"tool/call","seq":21,"time":0,"data":{"turn":1,"step":2,"callId":"call_goal_create","name":"create_goal","arguments":"{\"objective\":\"Finish the headless goal-tool snapshot proof\",\"max_goal_rounds\":7}"}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"tool/result","seq":22,"time":0,"data":{"turn":1,"step":2,"callId":"call_goal_create","content":[{"type":"text","text":"{\"goal\":{\"id\":\"goal-{{sessionId}}\",\"revision\":1,\"objective\":\"Finish the headless goal-tool snapshot proof\",\"phase\":\"active\",\"roundsStarted\":0,\"maxGoalRounds\":7},\"activation\":\"armed\"}"}],"isError":false},"sourceEventSeqs":[21],"surfaceOp":"append"}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"user/message","seq":23,"time":0,"data":{"content":[{"type":"text","text":"{\"goal\":{\"id\":\"goal-{{sessionId}}\",\"revision\":1,\"objective\":\"Finish the headless goal-tool snapshot proof\",\"phase\":\"active\",\"maxGoalRounds\":7},\"roundsStarted\":0,\"createdAt\":0,\"updatedAt\":0}"}],"source":{"kind":"goal","goalId":"goal-{{sessionId}}","revision":1,"round":0},"meta":{"kind":"goal/change","version":1,"operation":"create","goal":{"id":"goal-{{sessionId}}","revision":1,"objective":"Finish the headless goal-tool snapshot proof","phase":"active","maxGoalRounds":7},"roundsStarted":0,"createdAt":0,"updatedAt":0}},"surfaceOp":"append"}} {"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"step/end","seq":24,"time":0,"data":{"turn":1,"step":2}}} {"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"step/start","seq":25,"time":0,"data":{"turn":1,"step":3}}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":26,"time":0,"data":{"turn":1,"step":3,"chunk":{"type":"block-start","index":0,"blockType":"text"}}}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":27,"time":0,"data":{"turn":1,"step":3,"chunk":{"type":"text-delta","index":0,"text":"GOAL READY"}}}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":28,"time":0,"data":{"turn":1,"step":3,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"GOAL READY"}}}}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":29,"time":0,"data":{"turn":1,"step":3,"chunk":{"type":"usage","usage":{"inputTokens":35,"outputTokens":2}}}}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":30,"time":0,"data":{"turn":1,"step":3,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/message","seq":31,"time":0,"data":{"turn":1,"step":3,"content":[{"type":"text","text":"GOAL READY"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":35,"outputTokens":2}},"sourceEventSeqs":[26,27,28,29,30],"surfaceOp":"append"}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"step/end","seq":32,"time":0,"data":{"turn":1,"step":3}}} -{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"turn/end","seq":33,"time":0,"data":{"turn":1,"reason":{"kind":"completed"}}}} -{"type":"result","success":true,"sessionId":"{{sessionId}}","turn":1,"result":"GOAL READY","reason":{"kind":"completed"},"usage":{"inputTokens":85,"outputTokens":14}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":26,"time":0,"data":{"turn":1,"step":3,"chunk":{"type":"block-start","index":0,"blockType":"tool-call"}}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":27,"time":0,"data":{"turn":1,"step":3,"chunk":{"type":"tool-call-delta","index":0,"id":"call_goal_get","name":"get_goal","argumentsDelta":"{}"}}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":28,"time":0,"data":{"turn":1,"step":3,"chunk":{"type":"block-end","index":0,"block":{"type":"tool-call","id":"call_goal_get","name":"get_goal","arguments":"{}"}}}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":29,"time":0,"data":{"turn":1,"step":3,"chunk":{"type":"usage","usage":{"inputTokens":30,"outputTokens":4}}}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":30,"time":0,"data":{"turn":1,"step":3,"chunk":{"type":"finish","reason":{"kind":"tool-calls"}}}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/message","seq":31,"time":0,"data":{"turn":1,"step":3,"content":[{"type":"tool-call","id":"call_goal_get","name":"get_goal","arguments":"{}"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":30,"outputTokens":4}},"sourceEventSeqs":[26,27,28,29,30],"surfaceOp":"append"}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"tool/call","seq":32,"time":0,"data":{"turn":1,"step":3,"callId":"call_goal_get","name":"get_goal","arguments":"{}"}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"tool/result","seq":33,"time":0,"data":{"turn":1,"step":3,"callId":"call_goal_get","content":[{"type":"text","text":"{\"goal\":{\"id\":\"goal-{{sessionId}}\",\"revision\":1,\"objective\":\"Finish the headless goal-tool snapshot proof\",\"phase\":\"active\",\"roundsStarted\":0,\"maxGoalRounds\":7},\"activation\":\"armed\"}"}],"isError":false},"sourceEventSeqs":[32],"surfaceOp":"append"}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"step/end","seq":34,"time":0,"data":{"turn":1,"step":3}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"step/start","seq":35,"time":0,"data":{"turn":1,"step":4}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":36,"time":0,"data":{"turn":1,"step":4,"chunk":{"type":"block-start","index":0,"blockType":"text"}}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":37,"time":0,"data":{"turn":1,"step":4,"chunk":{"type":"text-delta","index":0,"text":"GOAL READY"}}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":38,"time":0,"data":{"turn":1,"step":4,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"GOAL READY"}}}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":39,"time":0,"data":{"turn":1,"step":4,"chunk":{"type":"usage","usage":{"inputTokens":35,"outputTokens":2}}}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/chunk","seq":40,"time":0,"data":{"turn":1,"step":4,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"assistant/message","seq":41,"time":0,"data":{"turn":1,"step":4,"content":[{"type":"text","text":"GOAL READY"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-flash"},"usage":{"inputTokens":35,"outputTokens":2}},"sourceEventSeqs":[36,37,38,39,40],"surfaceOp":"append"}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"step/end","seq":42,"time":0,"data":{"turn":1,"step":4}}} +{"type":"session_event","sessionId":"{{sessionId}}","event":{"type":"turn/end","seq":43,"time":0,"data":{"turn":1,"reason":{"kind":"completed"}}}} +{"type":"result","success":true,"sessionId":"{{sessionId}}","turn":1,"result":"GOAL READY","reason":{"kind":"completed"},"usage":{"inputTokens":100,"outputTokens":20}} diff --git a/packages/goal/tool-goal/README.md b/packages/goal/tool-goal/README.md index 6e0b25c567..346ac07dcc 100644 --- a/packages/goal/tool-goal/README.md +++ b/packages/goal/tool-goal/README.md @@ -6,9 +6,9 @@ The model-facing control surface for [`ctx.goals`](../goal/README.md): `get_goal - `get_goal()` returns the current goal or `null`, including the compare-and-set id/revision, durable phase, admitted/capped goal rounds, any blocker reason, and current process-local activation. - `create_goal(objective, max_goal_rounds?)` creates one goal from a direct top-level human turn. The model may infer long-running goal intent without an exact command phrase; non-human turns and subagents are rejected at execution. -- `update_goal(goal_id, revision, action, objective?, max_goal_rounds?, blocked_reason?)` supports `edit`, `pause`, `resume`, `complete`, and `blocked`. Replacements belong only to `edit`; `blocked_reason` is required only for `blocked` and is persisted with the stable code `model-reported`. +- `update_goal(goal_id, revision, action, objective?, max_goal_rounds?, blocked_reason?)` supports `edit`, `pause`, `resume`, `complete`, and `blocked`. Replacements belong only to `edit`; `blocked_reason` is required only for `blocked` and is persisted with the stable code `model-reported`. Strict-schema empty-string and zero fillers count as omitted, while meaningful values remain limited to their action. -All calls are exclusive, so a model-ordered batch observes earlier mutations and their new revisions. ACP and other clients receive pure generic cards: read for `get_goal`, other for mutations. +All calls are exclusive, so a model-ordered batch observes earlier mutations and their new revisions. ACP and other clients receive pure generic cards: read for `get_goal`, other for mutations. Mutation cards select the first meaningful action value and otherwise show the goal id, so accepted fillers never produce blank input. All three canonical values match the compact JSON already rendered to Native callers: `{ goal: null }` or `{ goal: { id, revision, objective, phase, roundsStarted, maxGoalRounds, blockedReason? }, activation }`. Programmatic consumers therefore receive the same domain structure without parsing the rendered JSON. diff --git a/packages/goal/tool-goal/src/index.ts b/packages/goal/tool-goal/src/index.ts index 96e6389f85..09953041fa 100644 --- a/packages/goal/tool-goal/src/index.ts +++ b/packages/goal/tool-goal/src/index.ts @@ -132,13 +132,13 @@ function resolveConfig(config: Config): ResolvedConfig { return { blockedAfterConsecutiveRounds: blockedAfter } } -/** Whether an optional string carries a meaningful action-specific value. */ -function hasText(value: string | undefined): boolean { +/** Whether optional text is meaningful rather than a strict-schema empty filler. */ +function hasText(value: string | undefined): value is string { return value !== undefined && value !== '' } -/** Whether an optional round cap carries a meaningful action-specific value. */ -function hasRoundCap(value: number | undefined): boolean { +/** Whether an optional round cap is meaningful rather than a strict-schema zero filler. */ +function hasRoundCap(value: number | undefined): value is number { return value !== undefined && value !== 0 } @@ -281,13 +281,11 @@ export function apply(ctx: Context, config: Config): void { const execution = goalToolExecution(ctx, exec) const ref = goalRef(args.goal_id, args.revision) const replacements = { - ...args.objective === undefined ? {} : { objective: args.objective }, - ...args.max_goal_rounds === undefined ? {} : { maxGoalRounds: args.max_goal_rounds }, + ...hasText(args.objective) ? { objective: args.objective } : {}, + ...hasRoundCap(args.max_goal_rounds) ? { maxGoalRounds: args.max_goal_rounds } : {}, } if (args.action === 'edit') { requireDirectHuman(ctx, execution) - // Some strict-schema providers emit empty placeholders for every declared - // optional field. Reject only values that could change another action. if (hasText(args.blocked_reason)) { throw new HarnessError('blocked_reason is valid only with action blocked', 'GOAL_TOOL_INVALID_UPDATE') } @@ -343,7 +341,11 @@ export function apply(ctx: Context, config: Config): void { presentCall: args => present( `${args.action === 'blocked' ? 'Mark' : args.action.charAt(0).toUpperCase() + args.action.slice(1)} goal`, 'other', - args.blocked_reason ?? args.objective ?? args.goal_id, + hasText(args.blocked_reason) + ? args.blocked_reason + : hasText(args.objective) + ? args.objective + : hasRoundCap(args.max_goal_rounds) ? args.max_goal_rounds : args.goal_id, ), })) } diff --git a/packages/goal/tool-goal/tests/tool-goal.spec.ts b/packages/goal/tool-goal/tests/tool-goal.spec.ts index 8823a6c00a..eaab21e202 100644 --- a/packages/goal/tool-goal/tests/tool-goal.spec.ts +++ b/packages/goal/tool-goal/tests/tool-goal.spec.ts @@ -145,8 +145,17 @@ describe('goal tool registration and presentation', () => { expect(ctx.tools.get('update_goal')?.presentCall?.({ goal_id: 'goal-1', revision: 2, action: 'blocked', blocked_reason: 'Waiting for a human choice.', })).toEqual({ card: 'generic', title: 'Mark goal', kind: 'other', rawInput: 'Waiting for a human choice.' }) + expect(ctx.tools.get('update_goal')?.presentCall?.({ + goal_id: 'goal-1', revision: 2, action: 'edit', + objective: 'ship', max_goal_rounds: 0, blocked_reason: '', + })).toEqual({ card: 'generic', title: 'Edit goal', kind: 'other', rawInput: 'ship' }) + expect(ctx.tools.get('update_goal')?.presentCall?.({ + goal_id: 'goal-1', revision: 2, action: 'edit', + objective: '', max_goal_rounds: 8, blocked_reason: '', + })).toEqual({ card: 'generic', title: 'Edit goal', kind: 'other', rawInput: 8 }) expect(ctx.tools.get('update_goal')?.presentCall?.({ goal_id: 'goal-1', revision: 2, action: 'resume', + objective: '', max_goal_rounds: 0, blocked_reason: '', })).toEqual({ card: 'generic', title: 'Resume goal', kind: 'other', rawInput: 'goal-1' }) expect(ctx.tools.get('update_goal')?.presentCall?.({ wrong: true })).toBeUndefined() }) @@ -438,11 +447,23 @@ describe('goal tool state transitions', () => { revision: goal.revision, action: 'edit', objective: 'edited', + max_goal_rounds: 0, blocked_reason: '', }, root.agent) expect(resultGoal(edited)).toMatchObject({ objective: 'edited' }) goal = ctx.goals.get(root.agent)! + const capped = await execute(ctx, 'update_goal', { + goal_id: goal.id, + revision: goal.revision, + action: 'edit', + objective: '', + max_goal_rounds: 8, + blocked_reason: '', + }, root.agent) + expect(resultGoal(capped)).toMatchObject({ objective: 'edited', maxGoalRounds: 8 }) + goal = ctx.goals.get(root.agent)! + const paused = await execute(ctx, 'update_goal', { goal_id: goal.id, revision: goal.revision,