Clicking "fetch available models" on a built-in provider went to the
network. That is the wrong source: pi-ai's registry is the authoritative
list for its own providers, and it carries the context windows and output
caps a `GET /models` listing does not disclose. Asking api.deepseek.com
what DeepSeek serves is both slower and worse, and against an endpoint
that answers a different shape it failed outright.
Interrogation is still keyed by settings namespace — the provider being
added has no route — but the request may now name the route it is
editing. An adapter that already describes that route answers from what
it knows, needs no endpoint at all, and never touches the network; only a
route the catalog does not describe reaches the wire, and one naming no
endpoint is told to set one or enter its models by hand.
`ConfigurableProviderView` gained `supportsDiscovery` so a surface offers
the action where a namespace can answer instead of hardcoding an adapter
family.
Three narrower corrections ride along. Discovery no longer claims Azure
or Codex: Azure authenticates with an `api-key` header and an
`api-version` query despite its OpenAI lineage, and Codex uses OAuth, so
both reported an authentication failure as a provider with no models.
Cancellation during the body read escaped as the raw abort reason rather
than a coded ABORTED. And the schema comment claiming the probe key is
never logged overstated it: the host neither stores nor returns it, but
it rides the client's outgoing envelope like every other secret-bearing
payload, and redacting that tap is a configuration-plane-wide change.
Once a pi-ai route became a declaration rather than a catalog lookup,
adding an OpenAI-compatible gateway meant knowing its model ids up
front. Most such endpoints publish that list at `GET /models`, but no
seam operation could ask: every one is keyed by a registered provider
route, and the provider being added has no route, no stored profile,
and no stored credential — the endpoint and key are values in a form.
Interrogation is therefore keyed by settings namespace, which a
configuration surface already holds from the configurable-provider
directory. `registerModelDiscovery` offers it per namespace,
`discoverModels` asks, and the request carries the draft itself. The
reply is candidates, not a catalog: every field but the id is optional
because most listings disclose nothing else, and adopting one is a
settings write like any other. Nothing here reads or writes settings or
credentials, so `settings.yaml` still decides what a route serves.
`llm.discoverModels` carries the same draft over the wire. Its apiKey is
the third and last payload a secret may ride, and it is never stored,
logged, or echoed; every refusal folds into `model-discovery-failed`,
naming the endpoint asked but never the credential offered.
The pi-ai side is a plain GET for OpenAI-compatible protocols only —
their listing shape is the one gateways, self-hosted servers, and the
official endpoints agree on. Others say so, sending the user to
hand-entry rather than reporting a guessed shape as an empty provider.
The reply is read under a four-megabyte ceiling held on the bytes
actually received, because the endpoint is a URL the user typed.
A documentation rescan on master moved this README's Testing section out
of the package. This branch was still editing that section, so the
rebase asked which structure wins; master's does, and the branch keeps
only the Catalog resolution section its own change adds.
Re-records the pair fingerprint against the merged text.
pi-ai resolves a request's apiKey override only through a provider that
declares an api-key method: resolveProviderAuth short-circuits to that
method when the override is present, and otherwise falls through to the
credential store and then to ambient discovery. A provider with no
api-key method at all therefore resolves to nothing, and the request
fails with "Provider is not configured" before any network I/O.
Two routes hit that. openai-codex ships OAuth alone, so moving off the
/compat dispatch broke a profile that names a key for it — the old path
handed the token straight to the provider. And a catalog route naming an
api was being rebuilt with the harness's own auth, so `openai: {api:
openai-completions}` stopped reading OPENAI_API_KEY, contradicting the
documented promise that omitting a credential keeps provider-native
discovery.
Auth is now one decision for both constructions. A catalog route keeps
its installed provider's auth, through an api override too: which
environment a provider reads belongs to the provider, not to the wire
format its models speak. A catalog provider with no api-key method gets
the harness method beside its own, but only when the profile names a
credential — a keyless codex profile keeps the honest refusal, since
this adapter holds no OAuth store to resolve through.
Materialization now spreads the installed entry instead of enumerating
the result, so a Model field this package does not model survives a
pi-ai upgrade; headers went missing from an nvidia route exactly that
way once already. providerInfo reports the configured displayName, which
also joins the registration facts so a rename re-registers rather than
leaving the old label in every selector. A refused registration swap
gets its own diagnostic naming the route, matching the directory swap
beside it.
The README documented endpoint interrogation this layer does not
implement, and still described unknown providers as kept-last-good after
they became legal declarations refused at the write point. The Agent
Note claimed per-model reasoning configurability the schema never had,
required capacities the route now defaults, and stated an apiKey
override that short-circuits unconditionally.
The adapter omits the seam's reasoning field whenever a model carries no
reasoning metadata, which is the model's own property and says nothing
about where the model came from. Both the JSDoc and the Agent Note read
as though only hand-declared models were meant, so a reader would infer
that the 251 installed-catalog models pi-ai marks as non-reasoning still
offer their single off level. They do not, and that is the point: a
picker holding only off misrepresents a provider that thinks by default,
because off dispatches the same bytes as naming no effort at all.
Behavior is unchanged; only the prose that describes it was narrower
than the contract. adapter.spec.ts already pins the catalog case through
openai/gpt-4.1 and catalog.spec.ts pins the hand-declared one.
Three defects surfaced while driving the Models page.
A hand-declared model needed an explicit contextWindow and maxTokens,
but a provider listing usually returns ids and nothing else — so the
page happily wrote a profile the adapter then rejected, which took the
whole namespace down silently. Capacities now fall back to the route's
`defaultContextWindow` (262,144) and `defaultMaxTokens` (32,768). Both
are guesses by construction, which is why they are route fields a
deployment corrects once rather than constants buried in the adapter;
the fallback sizes the model and never becomes a per-request cap.
That silent failure was the second defect. A schema-valid profile the
adapter could not serve was stored and only rejected later, disabling
every route in the namespace with nothing said. `dsh-settings` gains an
optional `validate` on registration — a check for what a schema cannot
express — and `llm-pi-ai` refuses an unserviceable section at the write
that produced it. A stored section that fails keeps the namespace's last
good value, as a schema failure already did, so an externally edited
document still cannot strand the owner. The plugin's own last-good
fallback goes with it: nothing reaching it can fail any more.
Third, a model with no reasoning metadata advertised the single level
`off`, which pi-ai translates to *omitting* the reasoning option — the
same request naming no effort produces. Selecting it disabled nothing,
so a provider whose default is to think kept thinking with `off` shown
as selected. Such a model now reports no reasoning capability at all,
which is the seam's way of saying the control is unavailable, and the
per-model `reasoning` flag is gone: without a thinkingLevelMap to spell
levels it could only invent them.
The protocol table narrows to the three a hand-declared route reaches
today, most-reached first so a surface offering a choice defaults to the
one gateways actually speak.
Review found four defects in the declared-provider work.
`PiAiAdapter` reused one `Models` collection and mutated it whenever the
configuration changed. `Models.streamSimple()` resolves its provider
lazily — when the stream is first consumed, which is after the adapter
awaits the route's credential — so a configuration change landing in
that window let an in-flight request finish under a configuration it
never resolved against, or fail on a provider that no longer existed.
Each resolution now produces an immutable snapshot and every operation
captures one before its first await, which is what makes the seam's
per-step freeze (`llm.prepareCall()`) hold end to end: switching models
mid-reply takes effect on the next step, never inside the one in flight.
`defaultMaxTokens` was materialized from the catalog's `Model.maxTokens`.
The two answer different questions: pi-ai requires that field as the
model's output capability, while the seam's is a cap the deployment
chose to send on requests naming none, so every request had started
carrying a number nobody picked. Only an explicitly configured cap
reaches the seam now.
The configurable-provider directory was refreshed by disposing its
registration and making a new one. A candidate set the registry refuses
— a profile keyed `deepseek-official`, which llm-deepseek declares —
left the whole directory withdrawn and the Models page empty, silently,
because the settings callback contains the failure. The seam's
registration handle now carries `replace()` with the same
validate-first atomicity `registerAdapter` has.
The protocol table offered every pi-ai streaming API, including four
whose authentication a profile cannot express: Bedrock signs with SigV4
over AWS credentials and a region, Vertex needs a project, a location,
and ADC, Azure needs provider environment plus an api-version, and Codex
uses OAuth. Offering them handed back routes that cannot authenticate.
Catalog routes still reach them through their own provider.
A pi-ai route had to name an installed catalog provider, served that
catalog's models verbatim, and could override only the endpoint. An
OpenAI-compatible gateway, a self-hosted server, or a model newer than
the pinned pi-ai release was therefore unreachable, and a stale context
window could not be corrected without upgrading the package.
A route is now a declaration whose defaults come from the installed
catalog. `catalog.ts` merges that catalog under the profile's own model
entries, `provider.ts` builds the pi-ai Provider (reusing the catalog
provider when the route keeps its protocol, so implementations this
package cannot reconstruct keep working), and the adapter serves every
operation from one `createModels()` collection. That also retires the
`@earendil-works/pi-ai/compat` import, which pi-ai documents as a
temporary entry point it deletes with its ModelManager migration.
Credentials stay on the harness seam: the resolved key rides the request
as pi-ai's highest-priority auth override, so `Models` holds no
credential store and a named-but-missing reference still fails loud
instead of falling back to an unrelated ambient key.
A model's configured maxTokens now reaches the seam as defaultMaxTokens.
- docs: fix cross-class note links (../feature/...), mirror the membership
reuse criterion in README.zh.md and the connectWorkspace architecture
note (EN/ZH/sidecars re-recorded)
- test(client): stray + member blank at the same path — the scan skips the
stray and reuses the member
- test(web): re-registration after delete now mints a fresh blank session
(the old cwd-only reuse is gone), so the account is no longer empty;
assert the retained seeded session is still not re-adopted
The blank-session reuse scan in connectWorkspace matched on cwd alone, so a
live blank session the CLI/TUI birthed at the host cwd (never accounted to
any workspace) was hijacked when + was clicked on a workspace registered at
that path, opening a session the sidebar shows under Ungrouped instead of
the clicked workspace. Reuse now also requires membership in the workspace's
sessionIds (the host's own membership rule); cwd-only matches fall through
to session.create({workspaceId}).
Fixes#1647
The mocked process.send consumed vitest's own fork-pool IPC messages and immediately ran the post callback, whose disconnect() severed the test worker's channel (process.connected is true under forks) — the whole spec's results vanished and win32-dialog-bindings.ts/win32-dialog-worker.ts fell to near-zero coverage on CI. The mock now records without invoking the callback or disconnecting; the real close lifecycle stays with built-worker.e2e.ts. Verified under both the threads and forks pools.
The win32 tier is exactly the koffi IFileOpenDialog child process; any failure surfaces as-is. The pwsh -> Windows PowerShell 5.1 cascade, the shared WinForms script, and the triple-miss AggregateError are deleted: koffi is a packaged dependency whose availability the install guarantees, so no mechanism fallback exists (the browse backend remains the fallback at the composition level). The pwsh-first DPI picker-fix note is consolidated into a new simplification note recording the reversal.
The koffi IFileOpenDialog conversation runs in a spawned child process instead of a worker thread: the dialog is the child's first window, so Windows activates it without a foreground call, and a native fault stays contained to the child. The driver maps the child's message protocol onto a promise and services aborts by posting WM_CLOSE to the dialog thread's windows, killing the child when the close budget is exhausted. The built worker ships as lib/worker.cjs (the ./worker export) under plain node, and win32-dialog.spec.ts returns to the thread-safe pool.
The koffi redesign moved the dialog from a spawned child process (which
inherits a foreground-activation right from the spawning app) onto a worker
thread of the same process, so Windows shows the dialog without activating
it — it opens behind the app with a taskbar flash. The app has no native
HWND to hand the seam, so raise from the driver: on the 'showing' notice
(the worker posts it right before Show, before the dialog window exists),
attach this thread's input queue to the dialog thread's, call
SetForegroundWindow on its top-level window, and detach — retried on the
close cadence until the window appears, stopped on settle/abort/success,
never blocking the pick. Injectable seam mirrors closeThreadWindows; driver
tests pin the raise and its retry; the in-process note records the
mechanism (both languages, pairing re-recorded).
The driver's switch over Win32DialogWorkerMessage handled all three current
kinds but had no default, so a fourth kind added to the worker protocol
would compile cleanly and silently no-op — settle() never called and the
pick dangles until worker exit. Add the local assertNever backstop (the
command-compact shape; this package does not depend on dsh-llm for the
helper) and the return the error case needs to avoid falling into it.
Round-five review finding.
The round-four review's v6 pass found three factual gaps and the v5 pass two
nits. Correct them before merge:
- The in-process note claimed a koffi signature mistake is 'contained to
the worker thread' — worker_threads share the process, so a native access
violation takes down the whole Node process with no PowerShell fallback.
State the real blast radius and record the deferred pkg-VFS worker-spawn
arm in Consequences (both languages, pairing re-recorded).
- The 2026-07-27 picker note claimed unconditional 'per-monitor-v2 DPI';
PMv2-less hosts (Server 2016 / Win10 1607) cascade to per-monitor or
system-aware. Say 'the best thread DPI awareness the host accepts'
(both languages, pairing re-recorded).
- built-worker.e2e.ts was not in any keyless gate (vitest.e2e config is
not part of the default unit run and builtBinSmokeGate's explicit list
missed it), so lib/worker.cjs load regressions passed keyless CI. Add it
to builtBinSmokeGate alongside the workflow-workerthread sibling.
- Remove the dead trailing 'void bindings' in win32-dialog-logic.spec.ts
and give native-picker.spec.ts the sibling module header it lacked.
The @module header still described the Windows adapter as the pre-PR 'STA
PowerShell FolderBrowserDialog' while the README and Agent Notes document
the koffi IFileOpenDialog primary with the PowerShell chain as fallback;
mirror the README's platform summary.
- setThreadDpiAwareness checks SetThreadDpiAwarenessContext's return value
and cascades per-monitor-v2 -> per-monitor -> system-aware; DPI stays a
deliberate cosmetic best-effort - a host accepting none (or lacking the
API, pre-1607) still gets the modern dialog instead of a downgrade to the
legacy fallback chain over a cosmetic concern.
- The mocked-koffi world now uses a distinctive 4-byte pointer width and
rejects mis-sized out-buffers and mis-divided vtable offsets, so a
regression to hardcoded 8s fails the suite (the ia32 bug class).
- A keyless built-worker e2e guard loads lib/worker.cjs under plain
worker_threads on POSIX (the workflow-workerthread shape).
- The 'loaded lazily' module claims are reworded to attribute laziness to
the dynamic import('koffi') calls, and the discarded close-attempt
rejection is named at its catch.
Review round two on the in-process dialog:
- Vtable slots and out-pointers use koffi.sizeof('void *') instead of a
hardcoded 8 - win32-ia32 (which Node and koffi both ship) would have read
method pointers from the wrong address and crashed in-process before any
fallback could run.
- runFolderDialog pairs every successful (incl. S_FALSE) CoInitializeEx
with CoUninitialize in the outermost finally, releasing the dialog first;
a failed init is deliberately unpaired. Pinned across fake-bindings and
mocked-koffi suites.
- The abort close budget starts unconditionally: a worker hung before the
showing notice (koffi import or COM init) now ends in terminate instead
of a dangling promise; WM_CLOSE posting still waits for the thread id.
- A triple miss (dialog + pwsh + 5.1) surfaces an AggregateError carrying
all three causes - the in-process tier's reason was previously
unrecoverable from the final PowerShell error.
- The stray '=>{ ' formatter artifacts are normalized to real blocks.
Both stale note claims from the review are fixed: the DPI note's
Consequences no longer claims an ENOENT classification or zero new
dependencies, and the 2026-07-27 picker note's Windows bullet now names
the in-process primary and keeps the PowerShell chain as fallback (both
languages, pairings re-recorded).
The workspace files constraint keys worker bundles on the ./worker export
(lib/worker.cjs, the workflow-workerthread shape); the descriptive source
entry stays win32-dialog-worker.ts and tsdown renames the bundle.
On a fast host the 1ms close budget can exhaust and reject between waitFor
ticks; a rejection with no listener yet counted as an unhandled error in the
Linux run.
The modern IFileOpenDialog becomes the primary win32 tier: a koffi-driven
COM conversation on a worker_threads worker (the modal Show never blocks
the host event loop), per-monitor-v2 DPI via SetThreadDpiAwarenessContext,
and abort service by re-posting WM_CLOSE to the dialog thread's windows,
with terminate+unref as the last resort (Node cannot interrupt a thread
blocked in native code, and such a worker must never hold the process open).
The PowerShell chain stays as the fallback tier with its trigger widened
from ENOENT to any pwsh failure, closing the review-flagged PowerShell 6
regression (no WinForms: exit 1, not ENOENT, so 5.1 never ran).
Layering keeps per-file coverage honest on every host: pure sequencing and
the driver test against fakes anywhere; the bindings run against a mocked
koffi COM world (the session-persistence-jsonl technique); POSIX hosts
drive the real spawn plumbing to its koffi-load rejection; win32 hosts run
a real open-and-abort-close smoke. The smoke joins processBoundTests: a
worker blocked in a native modal wedges the threads pool's teardown, while
a fork contains it. The worker bundles as its own CJS tsdown entry
(workflow-workerthread's pattern; no TLA), and the host module is imported
statically so the node-half bundle stays chunk-free.
Built-plane and real-COM behavior verified on native Windows: standalone
probes for the source worker, the built CJS worker, and the driver's abort
path all open and close the real dialog.
Agent Notes: new implemented/feature/2026-08-02-win32-in-process-folder-dialog
(bilingual) owns the decision; the DPI note is re-scoped to the fallback tier
it now describes and its AutoUpgradeEnabled attribution corrected (.NET Core
3.0 rewrote FolderBrowserDialog; the opt-out arrived in .NET 6).
.NET 10's modern FolderBrowserDialog renders Description as a bottom
strip above the folder input, and the 5.1 classic dialog as an unthemed
white box; the property is dropped entirely and a regression assertion
pins its absence.
Yichen Jiang
creatixchu
_Kerman
Turtle
imccyu
Wenlu Wang
imccyu
xjt
Huanqi Cao