mirror of
https://github.com/THU-MAIC/OpenMAIC.git
synced 2026-10-17 05:29:46 +08:00
@openmaic/editor@0.0.4
131
Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
 wyucandClaude Fable 5
|
8cc8845a2c |
feat(document): derived assets with lineage and an extraction cache (#1153 part 1) (#1164)
* feat(document): cache extraction as derived pool assets with lineage (#1153 part 1) * fix(document): harden the extraction cache keys and write path (#1153 part 1) * fix(document): bound alias trust and unblock the extract path from cache writes (#1153 part 1) * fix(document): keep every extraction-cache failure inside the degrade-to-miss net (#1153 part 1) Move the lookup's key construction (config fingerprinting included) inside its guarded region so a fingerprint failure degrades to a cache miss instead of rejecting the caller's extraction; guard the page-side dedupe fingerprint the same way (falls back to un-deduped extraction). Pin the remaining review coverage: malformed/missing createdAt treated stale, the documented future-dated residual, fingerprint-throw degradation on lookup/write/dedupe, and supersede leaving the superseded record's assets untouched. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(document): keep server-only extractor implementations out of the client bundle (#1153 part 1) * fix(document): fall back to the localized message for non-JSON extraction responses (#1153 part 1) A proxy error page (or any non-JSON body) on either the error or the success read now surfaces parseFailedMessage instead of a raw SyntaxError, making the docblock's fallback promise universal. Review nit from #1164. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
wyuc
|
e3fd494658 |
feat(tts): add Qwen TTS voice cloning (#1160)
* feat(tts): add Qwen TTS voice cloning via the registration adapter * feat(settings): add Qwen TTS clone-voice manager * fix(tts): harden Qwen voice cloning integration * fix(tts): enforce voice-bound Qwen synthesis * fix(tts): close Qwen voice clone tail issues * fix(tts): close final voice clone review issues * fix(agents): harden voice token parsing * fix(settings): hide the voice-clone model from manual selection * fix(agents): pin the narrator voice to the user's selected voice * fix(agents): keep narrator fallback alive when the pinned voice is unusable * fix(i18n): correct the narration fallback notice copy * fix(tts): bound the narrator voice retry to a single fallback hop * fix(tts): tolerate transient vendor errors in the voice existence precheck |
||
|
LING DUANandwyuc
|
a0dccbe78d |
feat(pi): add Native Child additive whiteboard tools (#1156)
* feat(pi): add Native Child additive whiteboard tools * chore(renderer): bump package version to 0.1.3 * fix(pi): reject invisible whiteboard mutations * fix(pi): close native whiteboard review blockers --------- Co-authored-by: wyuc <wang-yc24@mails.tsinghua.edu.cn> |
||
|
wyucandClaude Fable 5
|
5a3e81c4d9 |
feat(document): ingest uploaded sources into the asset pool and extract by id (#1153 part 0) (#1154)
* feat(document): declare a version on every extractor provider (#1153 part 0) * feat(api): accept asset-id input in the extract-document route (#1153 part 0) * feat(upload): ingest selected course materials into the asset pool (#1153 part 0) * fix(document): harden asset-id extraction inputs and close upload ingest leaks (#1153 part 0) * fix(document): scope extraction fallback and harden the asset-id path (#1153 part 0) - G1: byte fallback only on pre-extraction failures; PARSE_FAILED (422/500) is surfaced without re-running the paid extractor - G2: freeze the course-material set for the duration of generate-prep (guarded add/remove + disabled toolbar affordances, session built from the click-time snapshot, belt-and-braces drain) - G3: bound the pre-generation ingest await (~15s); timed-out ingests go byte-path and their late-resolving ids are released - G4: reject non-object JSON bodies (null/array/string/number) as 400 - G5: JSON-path provider pre-validation with generic messages, generic JSON-path PARSE_FAILED/422 responses, control-char-free log values - G6: size cap enforced via identify() before resolve() materializes bytes; too_large maps to the existing 413 * fix(document): mirror provider-hint semantics on the asset-id form and freeze prep-time controls (#1153 part 0) * fix(document): keep the asset-id form's responses free of caller input (#1153 part 0) Address review findings: answer the extractor-selection failure with a generic static message on the asset-id form (the registry's interpolated message carries the caller's MIME type; multipart keeps it verbatim, both sides now pinned by tests), dedupe same-batch material additions by content fingerprint inside the pure updater, document the network-error retry exception in shouldRetryWithByteUpload's docstring, and fix a test name asserting 500 but named 400. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
LING DUAN
|
2cbe4ae293 |
feat(pi): add Native Child whiteboard tools (#1152)
* feat(pi): add native whiteboard tools * style(pi): format native whiteboard tests * fix(pi): clarify whiteboard CAS handoff * fix(whiteboard): reconcile runtime projection on open |
||
|
LING DUANandwyuc
|
59adbef03b |
Simplify Pi agent tool ownership and completion (#1148)
Co-authored-by: wyuc <wang-yc24@mails.tsinghua.edu.cn> |
||
|
LING DUANandwyuc
|
0d308ed78d |
feat(pi): add Native Child Web Search (#1133)
* feat(pi): add native child web search * fix(pi): align native web search with provider system * fix(pi): harden native web search configuration and cancellation * fix(pi): guide Native Child web search selection --------- Co-authored-by: wyuc <wang-yc24@mails.tsinghua.edu.cn> |
||
|
LING DUANandwyuc
|
cdf949a017 |
feat(pi): add native child runtime with scoped Spotlight (#1111)
* feat(pi): add native child runtime with scoped Spotlight * fix(pi): address native child review feedback * fix(pi): suppress prefaced structured child output * fix(pi): restore native child text streaming --------- Co-authored-by: wyuc <wang-yc24@mails.tsinghua.edu.cn> |
||
|
wyucand杨慎
|
0a10af11cf |
feat(storage): opt-in indirect asset byte egress (#1007) (#1100)
* feat(storage): add opt-in indirect asset byte egress A deployment can now opt an asset byte GET into a 302 to a short-lived signed URL when the byte layer can sign. It is off by default, and byte-for-byte unchanged when off. The byte layer gains an optional signReadUrl capability; S3AssetByteStore implements it through the new optional @aws-sdk/s3-request-presigner peer, resolved lazily exactly like the client SDK, with a signer seam mirroring the commands seam so tests can bind doubles. The signed URL pins the contract's response headers -- the media type after the renderable allowlist, the fixed disposition, and the no-store cache posture -- via S3 response-header overrides. PgAssetStore mints the URL from the same ownership-checked transactional read a direct resolve performs, so authorization still runs per read before any URL exists, and a store whose byte layer cannot sign falls back to direct bytes. The contract documents the shape and the disclosure tradeoff: a Location into a hash-keyed byte layer names the content hash, so deployments that need the no-disclosure property keep direct egress. Refs #1007 * feat(app): wire ASSET_BYTE_EGRESS into the persistence route ASSET_BYTE_EGRESS=redirect opts asset byte GETs into the storage package's indirect egress; unset, direct, or an unrecognized value keeps the default byte-for-byte behavior, with a warning for the unrecognized case. The lazy byte-store wrapper now forwards signReadUrl and answers undefined when the resolved layer has no signer, so the PostgreSQL byte column degrades to direct bytes and the S3 layer signs without the wrapper knowing which it holds. The collector path is untouched: it builds its byte store through createAssetByteStore as before and only ever calls delete. Refs #1007 * fix(storage): let the packaged client survive redirect egress Cross-review round 1 on the indirect byte egress change found two problems. A byte response that followed a redirect carries the pinned content type but no X-Asset-Revision, and HttpAssetStore required both from the same response, so every cold resolve under redirect egress failed MALFORMED_RESPONSE. The client now detects the shape once -- a redirected byte response with no revision -- latches it, and probes HEAD before each cold GET from then on, taking the label from the probe. The probe predates the download on purpose: a replacement in between labels newer bytes with an older revision, which the next revalidation detects and corrects, while the reverse order could pin stale bytes under a fresh revision with no signal to repair it. The probe doubles as the miss check, so a redirect-mode miss costs no download. The direct path is byte-for-byte unchanged; the one double download per client lifetime on first contact is documented in the contract, which gains the client-side paragraph this behavior implements. signedUrlTtlSeconds also accepted lifetimes beyond the seven-day SigV4 presigning maximum, minting redirects the object store would reject. The option is now capped at construction. Refs #1007 * fix(storage): cap signed URL lifetime below the reclamation grace A signed URL whose lifetime exceeds the collector grace period can outlive its object: the last reference goes, the grace elapses, the collector deletes the object, and the still-valid URL errors at the object store. The previous ceiling -- the seven-day SigV4 maximum -- made that window days wide against the one-hour default grace. The cap is now fifteen minutes, and the contract text states both the ceiling and the rule for deployments that shorten their grace. Refs #1007 * fix(storage): decline signing when the optional presigner is absent An unresolvable @aws-sdk/s3-request-presigner made signReadUrl throw, which the registry surfaced as a failed asset read -- a 500 for a deployment whose only fault is a missing optional peer. The contract's answer for a byte layer that cannot sign is the capability fallback: return undefined and let the caller serve the bytes directly. Signing errors from a resolved signer still fail loud; only the missing capability declines. Refs #1007 * fix(storage): answer redirect egress as a descriptor to asking clients Following a 302 is not header-neutral: the platform fetch forwards the original request's headers to the object store's origin, stripping only Authorization, so a deployment whose credential travels in a custom header would hand it to the object store -- and the preflight the forwarded custom headers provoke commonly fails there besides. This replaces the latch-and-probe client shape with an explicit one: the client sends X-Asset-Egress: descriptor on every byte GET, a redirect-egress server answers 200 with a JSON { url, revision } body instead of a Location, and the client fetches the signed URL with no deployment headers at all. The revision comes from the descriptor, so the probe-first ordering and its one wasted download are gone too. The 302 remains the answer for consumers that did not ask; the descriptor response is marked by its own header so a JSON-media asset can never parse as one. Refs #1007 * fix(storage): negotiate the descriptor through Accept, not a custom header A custom X-Asset-Egress request header is not CORS-safelisted, so every byte GET to a cross-origin server became a preflighted request -- a regression for direct-egress deployments that never opted into anything, against servers with no reason to allow the header. The negotiation now rides Accept with a vendor media type, which is safelisted and costs no preflight, and the descriptor answer is identified by that media type as its Content-Type rather than by a marker header. Refs #1007 * fix(storage): sign outside the registry transaction and reserve the descriptor type Two review findings on the indirect egress path. resolveIndirect awaited the signer inside the coordinated read, but a signer on refreshable credentials can wait on the network, and a database connection plus the blob row's FOR SHARE lock would be held across that delay. The read now closes first -- the lock still puts the hash, label and revision in one snapshot -- and signing runs afterwards, which cannot observe anything the read did not. The descriptor media type is also reserved from renderableTypes: the client identifies a descriptor answer by that exact Content-Type, and an allowlisted asset served with it would be misread as a descriptor and fail MALFORMED_RESPONSE. The handler now rejects the configuration at construction, and the contract states the reservation. Refs #1007 * fix(storage): ship the signing deps and negotiate without ambiguity Three review findings on the indirect egress path. The app now declares @aws-sdk/client-s3 and @aws-sdk/s3-request-presigner as runtime dependencies, and the standalone build force-includes them for the persistence route: both are reached through deliberately untraced dynamic imports, so the shipped image could not resolve them, and a deployment opting into S3 or redirect egress would have found the capability silently absent. The descriptor negotiation matches exactly on both sides: the client compares the Content-Type essence, so a longer media type that merely begins with the reserved value still serves as bytes, and the server parses Accept media ranges, so a descriptor range sent with q=0 selects the redirect instead of the descriptor it explicitly rejects. Refs #1007 * fix(app): refuse redirect egress when the collection grace is shorter A signed URL must never outlive its object: with the handler's 60-second default lifetime, a deployment that sets ASSET_COLLECTION_GRACE_MS below ten minutes lets the collector delete an object while a URL minted against it is still valid. The route now refuses the combination at handler initialization, naming both variables -- a deterministic misconfiguration, loud at startup rather than silent at read time. Refs #1007 * fix(app): ship the signing SDKs without loading them Two findings on the standalone deployment of redirect egress. The outputFileTracingIncludes globs did not follow pnpm's scoped symlinks, so the standalone image carried neither AWS package and both S3 mode and redirect egress could not resolve their SDK at runtime. The packages are now server-external and referenced from literal -- but never called -- dynamic import thunks in the persistence byte-store wiring, which gets them traced into the image while module resolution still happens only on first S3 use. Verified against a local standalone build: node_modules/@aws-sdk now contains both packages, and the route tests that pin the never-resolve-unless-configured discipline still pass. Media type comparisons are also case-insensitive now, as HTTP requires, on both the Accept parse and the descriptor recognition. Refs #1007 * fix(app): treat an empty collection grace as unset Number('') is 0, so a present-but-empty ASSET_COLLECTION_GRACE_MS failed the redirect-egress coordination check and took persistence down with it, while the collector's own parsing reads the same value as the one-hour default. The check now trims and treats empty as unset, matching durationEnv. Refs #1007 * fix(storage): accept bytes too, omit signing on the known-PG wrapper, state the full tradeoff Three review findings. The client's descriptor request now advertises both representations -- the vendor descriptor type preferred, */* accepted -- so a strict content-negotiating layer can never answer 406 to a client that does in fact consume plain bytes. The lazy byte-store wrapper no longer advertises signReadUrl when no bucket is configured: the layer is known to be PostgreSQL at construction, and carrying the method made resolveIndirect take its ownership query and blob-row lock just to decline, then repeat them in resolve, on every cold GET. With a bucket configured the lazy validation is preserved and the wrapper still declines when the resolved layer cannot sign. And the contract's disclosure tradeoff now states the second edge: object-store GETs carry ETag and Last-Modified, response overrides cannot strip them, and because deduplicated PUTs rewrite the hash-keyed object, Last-Modified moves when another principal re-uploads the same bytes -- shared-object write timing, beyond byte equality. Deployments for whom that signal is sensitive must keep direct egress. Refs #1007 * fix(app): degrade on bad grace, share the ttl invariant, cover the real boundary The latest review round, four findings. A misconfigured ASSET_COLLECTION_GRACE_MS now degrades redirect egress to direct with a warning instead of failing the shared handler's initialization -- the asset backend is optional, and its misconfiguration must never take document and runtime traffic down. The package exports assertSignedUrlTtlWithinGrace so a deployment wiring the handler and the collector separately validates the invariant once at its own boundary: the handler caps the lifetime, but only the deployment knows both numbers. The contract states that nosniff does not survive the redirect -- response-header overrides cannot set it -- and why the client-minted blob makes that safe for conforming clients. And the route gains a real-boundary test: the Fetch<->Node adapter, the real createStorageHttpHandler, and the egress wiring, exercised with an actual byte GET through the composed stack instead of mocks. Refs #1007 * fix(storage): make the unsafe egress configuration unrepresentable Three findings on the indirect egress path, all resolved by moving a rule into a place where it cannot be bypassed rather than by adding a check a consumer has to remember. The grace/TTL invariant now lives in the option shape. byteEgress takes 'direct' or { mode: 'redirect', collectionGraceMs, signedUrlTtlSeconds? }, so enabling indirect egress without declaring the reclamation window it lives inside no longer type-checks, and the handler validates both numbers at construction. The flat signedUrlTtlSeconds option, its cross-field "requires byteEgress redirect" check, and the standalone 900-second cap as the only guard all go away; the ceiling stays, but alongside the ratio rather than in place of it. This is what the earlier exported assertSignedUrlTtlWithinGrace could not do -- it protected only consumers who called it -- and it costs nothing in compatibility because the option is new in this branch. The app resolves the grace through one shared parser with the collector, so the number the handler checks is the number the collector runs with. A signed URL whose object is gone is now a miss. The client maps a 404 from the byte fetch to ASSET_NOT_FOUND, matching the direct path: the entry was owned and readable when the URL was minted, so a reclaimed object is the same physical state the direct read reports as a miss, and what a read means must not depend on the deployment's egress setting. Signing still does not probe for the object -- that would restore the round trip this feature removes and make the mint's price vary with prior presence -- so the contract instead requires the byte layer to answer 404 for an absent object, and states that S3 needs s3:ListBucket to do so. Route coverage now spans the whole egress matrix through the real composed handler rather than only the direct case: 302 for a consumer that did not ask, descriptor for one that did, direct bytes when the byte layer declines to sign, and direct bytes when a short grace degraded the configured mode. Refs #1007 * feat(storage): forward byte egress options through reference server * fix(storage): fail closed on redirects and unconfirmed 404s in indirect byte egress A signed-fetch 404 now maps to a miss only when the object store's XML body declares NoSuchKey; every other 404 from the signed fetch fails loud, so a wrong bucket, access point, or endpoint can no longer make a live asset read as absent. The descriptor byte GET is sent with redirect: 'manual' and any 3xx answer is treated as an error, so a server that ignores the descriptor negotiation can never forward the deployment's custom credential headers to a redirect target. The ASSET_BYTE_EGRESS switch is now documented in .env.example and the server-persistence deployment section, together with the object-store CORS and s3:ListBucket prerequisites it requires, and the asset descriptor media type is exported from the package root and the asset/http subpath. --------- Co-authored-by: 杨慎 <117187635+cosarah@users.noreply.github.com> |
||
|
|
969574bee4 |
feat(storage): wire the asset backend into the app (#1007) (#1089)
* fix(persistence): stop corrupting binary response bodies in the route adapter The embedded persistence route bridges a Node `RequestListener` to the Fetch API by hand. Its response object is cast with `as unknown as ServerResponse`, so the compiler checks none of that surface, and two parts of it were wrong in ways only a byte-carrying handler would hit. `end` decoded any `Uint8Array` chunk with `Buffer.from(chunk).toString()`, which is UTF-8. `ServerResponse.end` accepts a `Uint8Array`, and those bytes are not necessarily valid UTF-8, so every unpaired byte became U+FFFD -- silent corruption with no error anywhere. The adapter now buffers chunks as bytes and builds the response from them. `write` was absent entirely, making any chunked handler a runtime TypeError rather than a compile error. It is part of the surface this object claims to implement, so it is now implemented. Neither reaches a caller today: the document and runtime handlers only ever end with JSON strings, which took the string branch. Both are traps laid for the first handler that carries bytes. Both cases are pinned by tests that fail before this change -- the binary one on a byte comparison, the chunked one with a 500 from the TypeError. The existing adapter round-trip test, whose comment already called this the most bug-prone code in the route, covered only the string path. * feat(storage): wire the asset backend into the app (#1007) The server asset registry shipped with #1007 but was reachable from nothing: the persistence route never passed an `assetStore`, and no file under `app/`, `lib/` or `components/` imported the asset server or client modules. This connects it, additively -- no DSL change, no version bump, and no behavioural change when server persistence is off. The route now builds a `PgAssetStore` on the same pool and transaction as the document store, with `PgAssetByteStore` by default and `S3AssetByteStore` when `ASSET_S3_BUCKET` opts a deployment into it. The AWS SDK stays an optional peer, reached only through a dynamic import on that branch. The offline collector is deliberately not scheduled here, and the route says so where a reader would look for it, because leaving it unscheduled grows storage without bound and scheduling it is the deployment's decision. On the browser side the asset pool gains a single-shot configuration seam modelled on the document store's, so that with server persistence enabled the pool is an `HttpAssetStore` carrying the same auth headers as the other stores. Three browser-specific mechanisms needed decisions rather than translation: `clearAssetPool` deletes an IndexedDB database today. In server mode there is nothing local to delete, and a "clear cache" action must not remove server-side assets -- that would destroy user data from a button promising to free space. It now revokes local object URLs and closes the client, and deletes nothing remote. The cross-tab replacement broadcast is unchanged. Its job is to make another tab drop a warm object URL when the bytes behind an id change, and that need is identical in server mode. The exclusivity proof that decides whether an in-place `replace` is safe reasons over this browser only: local documents, unflushed state, and a cross-tab presence probe. In server mode those cover a strict subset of the holders, since another device can reference the same id and no probe can see it -- and asking the server who else references an id would be an existence oracle over other principals. The proof therefore returns not exclusive unconditionally in server mode, and regeneration takes the existing fork path. That is the fail-closed direction and a graceful degradation rather than a break. The development authenticator had to grow the asset principal's required `key`; without it every asset request would have been correctly denied. It derives from the same learner partition, and a principal without one still gets no asset access. The route now states plainly that this key comes from a client-supplied header, so the cross-principal isolation the asset contract describes is not in force under this authenticator -- that warning previously lived only in the auth module's own header, where nobody mounting the route would see it. * fix(storage): close three defects in the asset app wiring (#1007) **The S3 branch could not load, and isolated nothing.** The route dynamically imported the storage package's S3 byte store, but that module statically imported the AWS SDK, so merely resolving it pulled the SDK into resolution and bundling. The route then separately performed an ignored native import of the SDK from the app's own scope, which fails with ERR_MODULE_NOT_FOUND because the SDK lives under the storage workspace package -- so a deployment that set the bucket got a broken store rather than an S3 one. There is now one lazy loader in one resolution scope, the storage module no longer imports the SDK at module load, and the bucket is validated at initialization rather than treated as valid because it is non-empty. A test mocks the SDK to throw on resolution and asserts that importing the byte-store module never resolves it. **A cross-tab replacement could leave a mounted lease stale.** The broadcast handler re-resolves, and `HttpAssetStore.resolve` coalesces concurrent calls for one id onto a single in-flight request -- correct on its own, and required by the contract. Together they meant a resolve triggered by the broadcast could merge into a GET that started before the peer's replacement, return the old revision, and leave the consumer on the old URL with nothing to dislodge it. The client gains an invalidation that retires its cached snapshot and advances the per-id generation without revoking any URL already issued; `release` would have been wrong here, since the contract keeps an issued URL valid until its holder releases it. Removing the invalidation call fails the new race test. **Clearing a server-backed pool left the closed client installed.** In server mode `clearAssetPool` closed the client and returned before clearing the module singleton, so every later `getAssetPool()` handed back a closed store until a page reload happened to intervene. Only the IndexedDB deletion is skipped in server mode now; the singleton is always cleared so the next call rebuilds through the configured factory. Also adds the tests whose absence let these through: a route round-trip carrying invalid UTF-8, the SDK-resolution assertion above, the broadcast race, and reopening the pool after a clear. The earlier tests were not wrong so much as aimed past the defect -- the route test used JSON text, and the S3 test injected a loader rather than exercising the imports. * chore(storage): bump package version to 0.3.0 * fix(persistence): make the route adapter faithful where it carries bytes The embedded persistence route bridges a Node `RequestListener` to the Fetch API by hand, and its response object is cast `as unknown as ServerResponse`, so the compiler checks none of that surface. Five divergences are closed here; the first two are the ones that corrupt data, the last three were raised in review. **`end` decoded bytes as text.** It ran `Buffer.from(chunk).toString()` on a `Uint8Array` chunk, which is UTF-8, so every byte outside that encoding became U+FFFD -- silent corruption with no error anywhere. The adapter now buffers chunks as bytes. **`write` was absent**, making any chunked handler a runtime `TypeError` rather than a compile error. **`write` callbacks ran inline.** Node invokes them after the chunk is handed off. A handler that writes its next chunk from each callback therefore recursed synchronously and could overflow the stack, and a throwing callback threw in the wrong execution phase. They are now deferred. **The encoding argument was ignored.** `write(s, 'latin1')` emitted `0xc3 0xa9` where Node emits `0xe9`, corrupting the body and potentially invalidating a `Content-Length` the handler had set. Both `write` and `end` now parse their full overload set and pass the encoding through. Node's own behaviour for an invalid encoding was established by probing `ServerResponse` on Node 22 rather than assumed: it throws `ERR_UNKNOWN_ENCODING` synchronously, so the adapter validates with `Buffer.isEncoding` and lets `Buffer.from` raise the native error. **Bodyless statuses were not suppressed.** Node discards writes after `writeHead(204)` or `writeHead(304)`; this passed the buffered bytes to the Fetch `Response` constructor, which throws for those statuses, so the request became a 500 instead of the intended 204 or 304. Buffered data is now discarded for 204, 205 and 304, and for `HEAD`. None of the five reaches a caller today -- the document and runtime handlers only ever end with JSON strings. They are traps laid for the first handler that carries bytes, which is why they surfaced while wiring the asset server backend, whose byte routes hit the first two immediately. Each is pinned by a test that fails when its fix is reverted. * fix(storage): keep { client, bucket } working for the S3 byte store `S3AssetByteStoreOptions.commands` was required and the store's methods called it directly, so the published `./asset/s3-bytes` export stopped accepting the `{ client, bucket }` construction it had always taken. Anyone not going through `loadS3AssetByteStore` either failed to compile or hit an undefined command factory at runtime. Make `commands` optional. Omitted, the store resolves the AWS SDK's own command constructors through a dynamic import on its first `write` / `read` / `delete` and caches them, so `{ client, bucket }` works again. Importing the module and constructing a store still never reach the optional peer dependency: a deployment that does not select S3 never resolves it. When the SDK cannot be resolved the call rejects naming `@aws-sdk/client-s3` rather than crashing on an undefined property, and the resolution is left uncached so installing the dependency fixes an already-constructed store. The lazily bound store is run against the byte-store contract, and the new isolation tests pin both that construction does not resolve the SDK and that an unresolvable SDK is reported by name. With the break gone this branch's change to the package is additive, so the version becomes a patch above main (0.2.4 -> 0.2.5) rather than the minor it was carrying. Refs #1007 Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * feat(persistence): run the asset collector in the shipped deployment `PgAssetStore.remove`, and a `replace` that changes content, only stamp `asset_blobs.unreferenced_at`. `AssetCollector.collect` is the only path that deletes anything, and the persistence route deliberately did not schedule it, on the grounds that scheduling is the deployment's decision. The deployment this repository ships is docker-compose.yml: the app and PostgreSQL, and nothing else that could make that decision. "The deployment decides" therefore meant "it never runs", and ordinary asset churn retained PostgreSQL bytes or S3 objects forever. Schedule it from instrumentation.ts, which Next runs once per server process — a route module can be instantiated more than once and has no shutdown hook, so a schedule started from one is really started per instantiation. Next 15 and later pick the file up with no config, so next.config.ts is unchanged. Collection is on by default with a 15-minute interval and the package's one-hour grace period, so the Compose stack is correct with no operator action. ASSET_COLLECTION_INTERVAL_MS, ASSET_COLLECTION_GRACE_MS, and ASSET_COLLECTION_ENABLED change or disable it, each documented where it is read. Without DATABASE_URL nothing is scheduled at all, and a failed pass is logged and retried on the next tick rather than ending the schedule or the process. Several instances may collect concurrently: each candidate blob row is re-checked and locked FOR UPDATE inside its own transaction, so collectors serialize on the row rather than racing. The comment says so, so nobody adds a distributed lock that is not needed. The byte-store selection moves to lib/persistence/asset-byte-store so the collector reclaims through the same layer the request path wrote through. A collector on the PostgreSQL byte store while the route wrote to S3 would drop the blob row and orphan the object permanently. Refs #1007 Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * feat(storage): bound each collector pass and report whether it filled An unbounded collect() is sized by however long a deployment ran before collection was scheduled: one statement selecting every eligible blob, then a transaction and a byte-layer delete each, in a loop nothing interrupts. The first pass over the Compose deployment that grew a backlog before collection was scheduled is exactly that pass. A pass now takes at most batchSize blobs (default one thousand) and returns. Ordinary churn between two scheduled passes is far below the cap, so a healthy deployment behaves as it did when a pass was unbounded; a capped pass costs the remainder one scheduling interval, which is what the interval is for. collect() still answers with the count, and the count alone cannot tell an empty backlog from a full batch -- a re-referenced or concurrently taken candidate is skipped, so even a full batch can collect less than batchSize. collectPass() returns the count together with capped, which is exactly "this batch was full, run again"; a caller draining a backlog loops while it is true. Candidates are taken oldest-unreferenced first, with the content hash as tiebreaker within one timestamp. Ordering by the hash alone would starve: digests are uniformly distributed, so a high-sorting blob waits behind every lower digest stamped after it, and under steady arrivals those keep coming. The tests pin the ordering against a planner that happens to answer the right rows without an ORDER BY. Refs #1007 Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix(persistence): keep asset backend faults off the shared handler Two review findings on the app wiring. A concrete asset pool instance is single-lifecycle, but resolveConfiguredAssetPoolStore handed the same object out again after clearAssetPool() had closed it, reinstalling a dead store as the live pool. The second handout now refuses loudly and points at the factory form, which rebuilds on every resolution. createPersistenceHandler eagerly awaited the asset byte store, so an invalid ASSET_S3_BUCKET or an unresolvable AWS SDK rejected the shared persistence handler and took document and runtime traffic down with an optional backend. Byte-store construction is now deferred to the first asset byte operation, and a failed construction is not cached, so the next asset request retries -- the route's own no-poisoned-singleton rule. Refs #1007 Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> --------- Co-authored-by: 杨慎 <117187635+cosarah@users.noreply.github.com> Co-authored-by: Claude Opus 5 <noreply@anthropic.com> |
||
|
PercyandPercy
|
b8bba7653e |
feat(courses): add folder grouping (#1005)
* feat(courses): add folder grouping behind a feature flag Group courses into user-created folders on the home page. Folders are device-local organization metadata kept in the existing IndexedDB database (new folders + course-to-folder mapping tables, Dexie v16); the course document aggregate owned by the DocumentStore is untouched. - Create / rename (inline) / delete folders (keep or remove members) - Move a course between folders via a hover menu - Navigate into a folder with a breadcrumb back to all courses - Search flattens the list, annotating each course with its folder - Deleting a course cleans up its folder membership - All UI strings internationalized across 9 locales - Gated behind NEXT_PUBLIC_ENABLE_COURSE_FOLDERS (default OFF) * test(courses): update db mocks and version assertion for folder tables The folder feature adds the `stageFolders` table to the deletion cascade and bumps the Dexie schema to v16. Update the affected test db mocks to include the new table and bump the version assertion accordingly. * fix(courses): stop course click-through when selecting from the move menu The move-to-folder menu is rendered inside the course card's clickable container. Selecting a folder item could let the click event reach the card's onClick (opening the course) because only the trigger button stopped propagation. Stop propagation on pointerdown for the trigger and on click for every menu item so selecting a destination never opens the course. * fix(courses): open a dialog for the move-menu new-folder entry The move-to-folder menu's inline "new folder" input never worked: a Radix DropdownMenu is modal, so a raw <input> inside it cannot keep focus — Radix closes the menu the instant the input is focused, dropping the field before anything can be typed. Move the new-folder entry out of the menu: it now asks the caller to open the existing NewFolderDialog and, on confirm, moves the requesting course into the freshly created folder. * feat(courses): drag-to-file folders and stacked cover thumbnails Per review feedback, make folders feel complete: - Drag a course card onto a folder tile to file it there. The tile turns into a clear drop target (ring + overlay) while a drag is over it. The hover 📂 menu remains as the accessible fallback for keyboard/touch. - Replace the placeholder folder icon with a stable stack of up to three member course covers (most recently updated frontmost). Empty folders keep the folder icon; the name and course count are always visible. * feat(courses): refine folder cover stack to a tidy fanned layout Tune the folder-tile cover stack per review: front cover centered and upright, rear covers peek out from alternating sides with a slight tilt and reduced opacity, soft shadow and hairline ring. Reads as a neat pile of course covers rather than an exaggerated fan. * fix(courses): address review — dialog mount, empty-list, rename validation, partial-delete refresh, breadcrumb dedup Per review (wyuc, CHANGES_REQUESTED): 1. Mount NewFolderDialog/DeleteFolderDialog outside the collapsible Recent subtree so they are reachable while it is collapsed; the New-folder button now expands the section before opening the dialog. 2. Keep the Recent/folder surface alive when the course list is empty so folders remain reachable and a first folder can be created. 3. Enforce folder-name validation (width + uniqueness) on rename as well as create, at the storage boundary (FolderNameError) and in the UI. 4. Refresh authoritative state on folder-delete both success and failure, so a partial "remove" failure does not leave stale cards/counts. 5. Drop the duplicated folder breadcrumb from the top section header; keep a single navigation breadcrumb in the content area. Add focused storage tests covering rename validation, membership writes, both deletion modes, and partial-failure propagation. * fix(courses): address 2nd review — folder-view layout, drop highlight, validation parity, a11y, i18n Per review (wyuc, CHANGES_REQUESTED on |
||
|
wyucand杨慎
|
1498ca52b2 |
fix(persistence): stop corrupting binary response bodies in the route adapter (#1088)
* fix(persistence): stop corrupting binary response bodies in the route adapter The embedded persistence route bridges a Node `RequestListener` to the Fetch API by hand. Its response object is cast with `as unknown as ServerResponse`, so the compiler checks none of that surface, and two parts of it were wrong in ways only a byte-carrying handler would hit. `end` decoded any `Uint8Array` chunk with `Buffer.from(chunk).toString()`, which is UTF-8. `ServerResponse.end` accepts a `Uint8Array`, and those bytes are not necessarily valid UTF-8, so every unpaired byte became U+FFFD -- silent corruption with no error anywhere. The adapter now buffers chunks as bytes and builds the response from them. `write` was absent entirely, making any chunked handler a runtime TypeError rather than a compile error. It is part of the surface this object claims to implement, so it is now implemented. Neither reaches a caller today: the document and runtime handlers only ever end with JSON strings, which took the string branch. Both are traps laid for the first handler that carries bytes. Both cases are pinned by tests that fail before this change -- the binary one on a byte comparison, the chunked one with a 500 from the TypeError. The existing adapter round-trip test, whose comment already called this the most bug-prone code in the route, covered only the string path. * fix(persistence): make the route adapter faithful where it carries bytes The embedded persistence route bridges a Node `RequestListener` to the Fetch API by hand, and its response object is cast `as unknown as ServerResponse`, so the compiler checks none of that surface. Five divergences are closed here; the first two are the ones that corrupt data, the last three were raised in review. **`end` decoded bytes as text.** It ran `Buffer.from(chunk).toString()` on a `Uint8Array` chunk, which is UTF-8, so every byte outside that encoding became U+FFFD -- silent corruption with no error anywhere. The adapter now buffers chunks as bytes. **`write` was absent**, making any chunked handler a runtime `TypeError` rather than a compile error. **`write` callbacks ran inline.** Node invokes them after the chunk is handed off. A handler that writes its next chunk from each callback therefore recursed synchronously and could overflow the stack, and a throwing callback threw in the wrong execution phase. They are now deferred. **The encoding argument was ignored.** `write(s, 'latin1')` emitted `0xc3 0xa9` where Node emits `0xe9`, corrupting the body and potentially invalidating a `Content-Length` the handler had set. Both `write` and `end` now parse their full overload set and pass the encoding through. Node's own behaviour for an invalid encoding was established by probing `ServerResponse` on Node 22 rather than assumed: it throws `ERR_UNKNOWN_ENCODING` synchronously, so the adapter validates with `Buffer.isEncoding` and lets `Buffer.from` raise the native error. **Bodyless statuses were not suppressed.** Node discards writes after `writeHead(204)` or `writeHead(304)`; this passed the buffered bytes to the Fetch `Response` constructor, which throws for those statuses, so the request became a 500 instead of the intended 204 or 304. Buffered data is now discarded for 204, 205 and 304, and for `HEAD`. None of the five reaches a caller today -- the document and runtime handlers only ever end with JSON strings. They are traps laid for the first handler that carries bytes, which is why they surfaced while wiring the asset server backend, whose byte routes hit the first two immediately. Each is pinned by a test that fails when its fix is reverted. --------- Co-authored-by: 杨慎 <117187635+cosarah@users.noreply.github.com> |
||
|
wyuc
|
3b4567076f |
refactor(generation): consume the package everywhere and delete lib/generation (#1090)
Part D of #1057 — the switch. All consumers import @openmaic/generation; lib/generation deleted (net -11,900 lines). PBL wiring preserved (single-call -> classified fallback -> injected loop), streaming outline route deduplicated onto buildOutlinePrompt, serverExternalPackages wired, and the change verified by a real-model four-kind end-to-end run against the production build with a rendered-classroom screenshot check. Closes #1057. |
||
|
|
fc66cc0603 |
feat(runtime): add learner whiteboard RuntimeStore foundation (#1075)
* feat(runtime): add whiteboard RuntimeStore foundation * Potential fix for pull request finding 'Comparison between inconvertible types' Co-authored-by: Copilot Autofix powered by AI <223894421+github-code-quality[bot]@users.noreply.github.com> * fix(runtime): reject invalid whiteboard appends before commit * fix(runtime): preserve special JSON keys during canonical clone * Potential fix for pull request finding 'Invalid prototype value' Co-authored-by: Copilot Autofix powered by AI <223894421+github-code-quality[bot]@users.noreply.github.com> * feat(runtime): add minimal learner whiteboard mutation --------- Co-authored-by: Copilot Autofix powered by AI <223894421+github-code-quality[bot]@users.noreply.github.com> Co-authored-by: wyuc <wang-yc24@mails.tsinghua.edu.cn> |
||
|
|
5d5acd6215 |
feat: claude search (#393)
* Added search tool support for Claude * New locales for i18n * Updated translations * Prevent default web search base URLs from being prepopulated * fixed coding style issues * fixed coding style issues * fixing lint errors * fix: add SSRF protection to fetchPageContent in claude.ts Agent-Logs-Url: https://github.com/joseph-mpo-yeti/OpenMAIC/sessions/099e53fb-e1db-49f4-af7f-4f17cb3d364a Co-authored-by: joseph-mpo-yeti <55380155+joseph-mpo-yeti@users.noreply.github.com> * fix: address all remaining review feedback from PR review thread Agent-Logs-Url: https://github.com/joseph-mpo-yeti/OpenMAIC/sessions/e1ab8fe2-fe0d-45a5-a0af-30945b99c268 Co-authored-by: joseph-mpo-yeti <55380155+joseph-mpo-yeti@users.noreply.github.com> * fix: rename ambiguous variable names in settings store for clarity Agent-Logs-Url: https://github.com/joseph-mpo-yeti/OpenMAIC/sessions/e1ab8fe2-fe0d-45a5-a0af-30945b99c268 Co-authored-by: joseph-mpo-yeti <55380155+joseph-mpo-yeti@users.noreply.github.com> * fixing lint errors * fix: address 5 items from second PR review thread Agent-Logs-Url: https://github.com/joseph-mpo-yeti/OpenMAIC/sessions/09469d80-8e91-449b-8bc2-9c11f83555df Co-authored-by: joseph-mpo-yeti <55380155+joseph-mpo-yeti@users.noreply.github.com> * fixing model verification error * fix: address 3 items from third PR review thread Agent-Logs-Url: https://github.com/joseph-mpo-yeti/OpenMAIC/sessions/0aa49d19-df5b-4a08-a7e1-2295ae9cfc38 Co-authored-by: joseph-mpo-yeti <55380155+joseph-mpo-yeti@users.noreply.github.com> * fixing lint errors * fix: resolved Copilot review comments * fix: returning string promise for ssrf-guard * fix: returning string promise for ssrf-guard * fix: updated locales, added allowed_callers for claude search tools, handle duplicated claude model keys * fix: updated locales, added allowed_callers for claude search tools, handle duplicated claude model keys * updated tests * using Anthropic SDK for search and connection test in search settings. * uses @ai-sdk with a custom fetch including allowed_callers in payload instead of adding Anthropic SDK as a direct dependency * fixed format * removed @anthropic-ai/sdk as direct dependency * fix: resolve TypeScript errors in claude web-search tests Cast mock.calls through unknown to avoid tuple-index type errors on Vitest mock objects whose parameter types are inferred as empty tuples. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * feat: add Claude as a web-search provider on the multi-provider framework Forward-port of the original Claude web-search work onto main's searchWeb dispatcher, per review guidance on #393: - lib/web-search/claude.ts: searchWithClaude via the AI SDK Anthropic provider and Anthropic's native web_search server tool, called through callLLM for usage accounting. Returns the model's answer plus cited sources deduplicated by URL. The page-content enrichment fetch from the original PR is removed entirely, resolving the P1 SSRF finding (no server-side fetches of attacker-influenced result URLs remain). - allowed_callers: ["direct"] is still injected at the fetch layer; the AI SDK's provider-defined web_search tools do not emit it. - Tool version picked from the model: web_search_20260209 for 4.6+ Opus/Sonnet, basic web_search_20250305 for Haiku/older; maxResults maps to the tool's max_uses. - Registry/dispatcher/route wiring: claude provider entry, claudeModelId passthrough (web-search route, classroom generation, generation preview), WEB_SEARCH_CLAUDE_* env config with server-pinned model via WEB_SEARCH_CLAUDE_MODELS, and client base-URL allowlisting for api.anthropic.com. - Settings: per-provider modelId with a model selector in the Claude panel; locale keys added to all ten languages. - Tests: new adapter suite plus claude cases in the dispatcher, constants, route, provider-config, web-search-config, and settings store suites. Closes #392 Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01E9KqAGmKCtBwMC4HtT4JKE * fix(web-search): trim whitespace around the Claude API key and base URL Keys are pasted by hand into Settings, and nothing in the client key path trimmed them: a trailing newline or space was sent verbatim as the x-api-key header value and came back from Anthropic as 'invalid x-api-key', with nothing in the message pointing at whitespace. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix(web-search): correct Claude adapter tool selection, base URL, and timing Addresses the four P2 review findings on #393: - Route dated Claude 4.0 ids (claude-sonnet-4-20250514, claude-opus-4-20250514) to the basic web_search_20250305 tool. The blocklist regex missed them, so they were sent web_search_20260209, which pre-4.6 models reject. Replaced with an allowlist of the models that have dynamic filtering, so unrecognized ids also fall back to the basic tool -- the safe direction, since 4.6+ models accept both. - Pin allowed_callers: ["direct"] only on web_search_20250305. Injecting it into every request disabled the code-execution caller that provides dynamic filtering on web_search_20260209. - Normalize the bare Anthropic root to /v1 before handing the base URL to the SDK, which appends /messages verbatim and would otherwise 404. The adapter is the single choke point for the client allowlist, the server env var, YAML, and the default; other hosts pass through untouched. - Return responseTime in seconds, matching every sibling adapter's WebSearchResult contract instead of exposing milliseconds for Claude. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com> |
||
|
wyuc
|
0ae21aa559 |
refactor(pbl): retire v1 write path and dead UI, converge on v2 (#1060)
* refactor(pbl): retire v1 write path and dead UI (#1058) Remove the legacy generator, renderer branches, chat API, and compatibility mirror writes. Keep stored v1 scenes readable through documented legacy adapters, and surface ordinary PBL planner double failures instead of degrading. * fix(pbl): contain planner failure to the scene, support legacy scenes in action regeneration, tighten legacy guards - Contain typed planner failures per scene while preserving retry and skip behavior. - Upgrade legacy PBL content before regenerating scene actions. - Apply video-export import boundaries to legacy readers. - Harden legacy rendering and empty-state handling. - Restore legacy upgrade compatibility coverage. - Remove retired translations and refresh fallback narration and legacy cover documentation. * fix(pbl): keep tolerant legacy configs renderable, guard invalid chat timestamps * chore(dsl): bump to 0.6.3 for README accuracy fix The README no longer references the deleted PBLProjectConfig symbol; the version gate requires a bump for any publishable package input change. * fix(pbl): harden legacy upgrade and persistence contract * fix(pbl): align action prompt with v2 workflow and validate project containers * fix(i18n): remove retired PBL v1 keys from the new fr-FR locale The fr-FR locale landed on main after this branch removed the v1 UI key families; prune the same 38 keys so all 10 locales align with en-US. * fix(generation): sync packaged pbl-actions template with the v2 wording The generation package that landed on main carries a byte-identical copy of the app's pbl-actions template; sync the v2 rewording and bump the package version. * fix(pbl): normalize legacy content and preserve planner status * fix(pbl): propagate provider status and ground action prompts * fix(pbl): classify planner failures for retry, default the project summary prompt input * fix(pbl): fall back to legacy data when v2 payloads are damaged, harden summary building * refactor(pbl): resolve hybrid content through a validity-aware resolver * fix(pbl): keep one v2 acceptance criterion across validator, resolver and persistence Persistence gates on the same container predicate as the validator and the renderer's resolver: payloads the renderer will not treat as v2 are inert bytes everywhere, avoiding runtime side effects for documents the write barrier would reject. An explicit null proficiencyAssessment stays processable. * fix(pbl): require runnable projects and harden legacy fallback * test(pbl): exercise the real save reload and hydration chain * fix(pbl): require an instructor and usable microtasks before v2 takes precedence * fix(pbl): fall back to the loop planner unless the provider failed or the user cancelled * fix(pbl): require role and task identities before v2 takes precedence Runtime normalization can repair status and create the instructor thread, but it cannot synthesize the role id that binds the thread, the microtask ids used by lookup and update paths, or the labels the workspace renders. The runnable predicate and its video-export mirror now require them, so identity-less payloads fall back to usable legacy data. |
||
|
|
ceb05adb41 |
feat(media): allocate generated assets through the registry (#1007 part 2, step b) (#1039)
* feat(media): establish shared asset ownership primitives Introduce global browser asset-pool ownership, asset-reference collection, stage reclamation planning, and lease-based URL access. Carry allocated media identity through storage and generation boundaries with regression coverage. * fix(media): enforce safe resolution across every consumer Route image, video, thumbnail, presentation, and video-export consumers through one resolution state machine. Prevent opaque allocated or generated references from reaching render and export sinks, with fallback and ownership tests. * fix(media): protect document-owned assets across mutations Allocate pool bytes before compatibility writes and document commits, then roll back uncommitted generations safely. Preserve document ownership across edits, retries, imports, speech generation, scene changes, and stage deletion. * fix(media): scope retries and tighten ownership guard Scope retries to the target scene and slide and refuse ambiguous shared-reference mutations. Expand retry rendering coverage and keep direct pool URL resolution behind the shared lease owner. * fix(storage): avoid nested lock during stage cleanup Execute prepared reclamation plans against an explicitly deleted document so the compatibility cascade cannot re-enter the per-document lock. Cover deletion of stage-owned media rows even when the document has no references. * refactor(media): confine reclamation to stage deletion Remove inline pool and compatibility-row cleanup from element, speech, scene, and audio replacement flows. Keep whole-stage reclamation behind explicit stage ownership, preserve stage-less legacy audio rows, and document deferred document-truth sweeping. * fix(media): close retry and resolution gaps Restore shared source tasks after successful forks, scope retries across both whiteboard locations, and wait for parallel TTS workers before rollback. Resolve background media through import, rendering, and PPTX export paths while keeping retry controls visible over last-good bytes. * test(media): execute the consumer safety matrix Replace source-substring checks with resolver seam execution across all six UI consumers, stage hydration, and both exporters. Pin each rollback layer independently and harden the ownership guard against aliased pool imports. * chore(packages): publish the additive DSL field Bump the DSL patch version for the optional speech-action field. Keep the transitional reclamation policy app-owned and document the legitimate transaction rollback removals there. * fix(media): make retry rollback task-safe Delay allocation task re-keying until final document reconciliation succeeds, restore shared source tasks on failed forks, and require exact stage-whiteboard targets before falling through from a missed scene. * fix(media): clear private assets and refresh leases Delete the asset-pool database during the confirmed local-data wipe. Notify the app lease layer after same-id replacement so mounted consumers re-resolve current bytes without reaching into storage internals. * test(media): pin closure safety guards Exercise the CSS allocation boundary, unfiltered legacy-row ownership, delete-and-undo byte survival, and real distinct consumer seams. Restore the video-only manifest overwrite condition and share the direct video resolution hook across both element variants. * test(media): narrow rollback element assertion Narrow the reconciled slide element to an image before checking its source so the rollback regression remains type-safe under the full root compiler configuration. * fix(media): close lease refresh races Gate the first batch publication by unique resolved refs, then publish every replacement snapshot without mutating the prior React state object. Serialize invalidation behind pending releases, evict rejected refreshes, and register replacement observation at the pool boundary. * fix(media): reopen pool after clear failures Always evict the singleton once its store has been closed, including blocked and failed database deletion paths. Report blocked deletion as deferred and prove a later write uses a fresh live store. * fix(media): isolate shared retry progress Track forked regeneration under the selected element until it receives a fresh asset identity, leaving the shared source task and bytes untouched. Surface targeted failures through renderer task lookup and skip the redundant fork reconciliation lock. * fix(media): close final asset retry gaps Keep blocked asset clears fail-loud until a successful retry, with actionable settings guidance. Clear failed shared-fork state across durable and live key spaces, and pin renderer lookups, lease publication identity, and committed-ref rollback protection. * fix: preserve actionable retry failures Localize the blocked cache-clear recovery hint across every supported locale and pin the deferred-error mapping. Retain durable fork failure rows while retries run, deleting them only after successful generation so unstructured failures survive reload. * fix(media): hydrate legacy stored video thumbnails Home-page recent-video thumbnails regressed for legacy Dexie mediaFiles rows keyed by gen_vid placeholders: the reworked hydration resolved the row's bytes through the sealed resolver but never surfaced the stored blob (and its poster) as object URLs for the preview card, so the CI recent-video-thumbnail e2e specs found no visible element. Hydration now materializes legacy stored video rows into blob URLs for both the element src and poster while keeping the resolver invariants: opaque refs still never reach a DOM src, and concrete addresses are never blanked. Unit pins cover the seam so the vitest suite catches this class without a browser. * fix(media): resolve sole restored legacy video Classroom playback restored tasks by exact document media references. Legacy gen_vid references can outlive the key used by the one persisted video row, leaving the player on a placeholder even though bytes were restored. Select the sole completed stage video only for legacy sequential refs after exact and reconciled matches. Keep exact failures authoritative, refuse ambiguous candidates, and cover success, ambiguity, and failure precedence in unit tests. * fix(media): scope legacy video recovery Decide restored legacy video recovery once from the complete document and record it through the shared task lookup consumed by playback, editing, and resolved slides. Keep ambiguous documents as placeholders, apply the same decision to thumbnail hydration, and preserve exact failure precedence. * fix(media): exclude claimed video recovery tasks Model restored legacy recovery as a two-pass match across document video elements and task rows. Remove tasks claimed by exact, targeted, or placeholder lookup before applying the sole-candidate fallback, and cover the ownership/cardinality matrix. * fix(media): unify video element resolution Centralize source, task, poster, and legacy recovery decisions for every video consumer. Ensure direct URLs win over opaque refs and play_video waits on element-targeted retry tasks. * refactor(media): route legacy recovery through resolver Let document-aware consumers request legacy video recovery through the unified element binding API, keeping thumbnail hydration on a single decision path. * fix(media): preserve import refs and prefer pool bytes Recognize unambiguous extensionless relative media addresses during classroom import. Resolve allocated export and thumbnail assets from the shared pool before falling back to lagging compatibility rows. * fix(export): preserve concrete video sources Route PPTX video elements through the shared media binding resolver so an unresolved opaque reference cannot replace a playable source. Complete browser and persisted-store cleanup when asset-pool deletion is deferred, while retaining distinct hard-failure behavior. * fix(media): fork retries without exclusive ownership Enumerate logical asset owners across every persisted document before allowing global pool replacement. Thread explicit targets through fresh-id rewrites and cover cross-document aliases plus unreadable ownership. * fix(media): guard global asset reclamation Share a fail-closed persisted-document liveness check between stage deletion and retry replacement. Preserve cross-document pool aliases while deleting stage-owned compatibility rows and cover enumeration failures. * fix(media): cover complete slide asset references Route slide media traversal through a shared mutable slot contract so backgrounds participate in export, thumbnail hydration, collection, and rewrite lifecycles. Snapshot complete surviving-document refs once per reclamation and preserve manifest-only owners while retaining fail-closed behavior. * fix(media): preserve exclusive retry asset ids Allow targeted retries to replace exclusively owned pool assets in place. Keep shared and unprovable ownership paths on fresh allocations, and pin production-shaped retries plus compatibility-row cleanup. * fix(media): revalidate asset bindings at completion Recheck repository-wide ownership before replacing generated media and fork scoped retries when exclusivity changed. Route video export selection through the unified resolver and keep concrete posters independent of task state. * fix(media): count unflushed owners and broadcast replacements The completion-time exclusivity proof read only the persisted document, but slide duplication updates the Zustand aggregate synchronously and schedules persistence behind a debounce. A retry finishing inside that window saw a single persisted owner and replaced the bytes behind a reference the duplicate also held. The proof now also counts owners in the live stage snapshot when that snapshot represents the stage being retried, so an unflushed duplicate forks instead. Same-id replacement notifications were realm-local, so a second tab showing the same classroom kept its lease pinned to the superseded blob URL. The notification now travels over a BroadcastChannel; each receiving realm runs its own observers against its own pool, so a spoofed message can at most force a re-resolve. A missing or failing channel never fails the replacement. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix(media): bind replacement listeners and spare shared audio rows A realm that only renders never sends a replacement, so binding the channel from the sender path left passive tabs deaf to peers. Binding now happens where the observer is registered, when the asset-pool module loads, and the receiving realm resolves its own pool lazily so a cleared or unavailable pool degrades to the next resolve instead of throwing. Observer notifications ran under Promise.all, so a rejection surfaced after BrowserAssetStore.replace had already committed and turned a durable success into a reported failure. They are settled individually now; the callback is wrapped because a synchronous throw would otherwise escape before allSettled sees the array. audioFiles rows are keyed globally by audioId, so deleting a stage removed the sole row for an id a surviving document still referenced — playback and both export paths read that table directly and cannot fall back to the preserved pool blob. Rows are now filtered against surviving references, while a failed enumeration still withholds only the irreversible pool removal. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix(media): serve replaced bytes from the pool everywhere Unknown survivor liveness deleted every planned audioFiles row. Those rows are keyed globally by audioId, and playback plus both export paths read the table directly, so losing one is as irreversible for them as removing the pool entry. Unknown liveness now preserves the rows too, leaving bounded garbage for a later pass that can prove exclusivity. The earlier pool-first change covered PPTX, video collection and thumbnail hydration but missed classroom ZIP export, which still serialized the stale compatibility row after a lagged same-id replacement, shipping media the classroom no longer renders. Auditing every direct reader of the media and audio tables surfaced the same gap in playback: speech regeneration also replaces bytes under a stable id and does not roll the pool back when the compatibility write fails, so the player kept serving superseded narration. It now resolves the pool first and falls back to stored rows for legacy and imported audio. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix(audio): replace exclusively owned speech clips in place regenerateSpeechAudio accepted the action's audioId but always passed undefined as replaceAssetId, so regenerating an exclusively owned pool-backed clip allocated a new asset every time, rewrote the action and orphaned the previous pool entry and compatibility row until stage reclamation — contradicting the stable-id path generateAndStoreTTS already implements for media. Ownership is now established before synthesis, and the rule itself moved to the shared reference module so media retries, poster replacement and speech regeneration consume one implementation instead of restating it. An exclusively owned clip keeps its id and has its bytes replaced; a shared clip, a legacy id with no pool entry, or unprovable ownership still gets a fresh allocation so other holders keep their audio. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * chore(media): drop the import left behind by the ownership move Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix(media): resolve audio pool-first and fence peer realms Stable-id TTS regeneration commits replaced narration to the pool before the audioFiles mirror write, so a failed mirror leaves the row stale. AudioPlayer already resolved pool-first, but audioObjectUrl, collectAudioFiles and the video timeline dependencies still read Dexie directly and would serve the superseded clip. All allocated-audio readers now share one resolver, with Dexie kept as the fallback for legacy and imported rows. The exclusivity proof modelled unflushed owners in the active realm only, so another tab duplicating the same asset during its save debounce could still be observed as a single owner and have its bytes replaced globally. A peer's pending state cannot be read across realms, so presence is probed instead: any realm holding the stage forces the fork path. A deferred asset-pool deletion no longer reloads the page. The database is still on disk and the guidance asks the user to close the other tab and retry, which the reload discarded. The decision moved into a shared helper so the rule is pinned rather than living inline in the component. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix(media): fail closed when presence cannot be probed The presence helper documented that an unanswerable probe must count as a peer, but every unavailable path — no BroadcastChannel, a constructor that threw, a send that failed, and the window before the pool's asynchronous binding completes — returned false, so the ownership proof cleared a single local owner and replaced globally shared bytes in place. Probing now returns present, absent or unknown, and only a probe that was actually sent and went unanswered is absent; the ownership decision treats unknown exactly like present. The pool declares its binding intent synchronously so a probe issued during the load-time window waits for the bind instead of concluding that presence is unavailable, and releases that gate if the import fails. Coverage reaches the write boundary: with presence unknown, a production-shaped targeted retry forks to a fresh id instead of calling replace, and the original bytes stay intact for a peer's unflushed owner. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * chore(dsl): bump to 0.6.3 after the release dedupe took 0.6.2 Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix(i18n): add the deferred-clear guidance to fr-FR The locale landed on main after this branch added the key, so the alignment check flagged it as the one missing translation. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5 <noreply@anthropic.com> Co-authored-by: 杨慎 <117187635+cosarah@users.noreply.github.com> |
||
|
wyuc
|
3f5c51bdba |
refactor(pbl-v2): split the planner core from the loop; inject the LLM call path (#1069)
Implements #1062. Pure planner-core extracted from the loop's file; both planners receive their LLM entry as a parameter (call sites pass the real callLLM unchanged); operations/ split into kernel vs runtime with a lint-enforced boundary; planner prompts pinned byte-identical via fixtures generated from the pre-refactor implementation. |
||
|
yipwingtimandwyuc
|
c556100999 |
fix: keep vocational mode toggle thumb within track (#1032)
Co-authored-by: wyuc <wang-yc24@mails.tsinghua.edu.cn> |
||
|
Nguyễn Lê TrườngandClaude Opus 5
|
36d6c14a3d |
fix(i18n): load every Inter subset so non-Latin text keeps the UI font (#1026)
`app/layout.tsx` pointed next/font at a single subset file, `inter-latin-wght-normal.woff2`. Only the stylesheets shipped by @fontsource carry the per-subset `unicode-range` declarations, so nothing requested the other subsets and every character outside Latin fell back to an arbitrary OS font — rendering in a different typeface, weight and metrics from the surrounding text, often mid-word. This already affects a shipped locale: `ru-RU` is Cyrillic, and `inter-cyrillic-wght-normal.woff2` was never loaded. `pt-BR` and `es-MX` lose the accented characters that live in `latin-ext`. Declaring the remaining subset files as sibling faces of the same family does not fix it. Faces with identical descriptors and no `unicode-range` are not tried per glyph — the browser picks one — and next/font/local has no way to express `unicode-range`. Importing the @fontsource stylesheet does, so that is what this switches to. `--font-sans` moves to `globals.css` because the family no longer comes from next/font's generated class. The typeface is unchanged: still Inter, now with all seven subsets it already shipped in node_modules. Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
|
f378c25a53 |
feat(storage): route settings/user-profile persist through KVStore (no legacy migration) (#1001)
* feat(storage): route zustand persist through KVStore The settings and user-profile stores persisted through zustand's default `localStorage` storage, while the rest of the app's small keyed values had already moved to `@openmaic/storage`'s `KVStore` — two unrelated mechanisms over the same browser API, which is the split-brain the storage RFC set out to remove. `kvPersistStorage` shipped with the package but had no importer. Add `lib/store/kv-persist.ts`, the app seam that adapts a `KVStore` into a zustand `PersistStorage`, and wire both stores to it under an explicit `account` scope. Store logic is unchanged: `version` / `migrate` / `merge` still belong to the stores, only the bytes' destination moves. Existing installs are adopted rather than orphaned. On first read the seam moves the raw `localStorage` entry into the KV scope — write first, then drop the raw key, so an interrupted migration leaves the value readable from one side or the other but never from neither. Once KV holds the entry it is authoritative, so a stale raw entry reappearing cannot roll back newer state, and re-running can neither duplicate nor resurrect anything. An entry that is not a readable persist envelope is left exactly where it is. Hydration is now asynchronous, because a `KVStore` may be remote. The browser backend resolves within a few microtasks of the store module being evaluated, so nothing user-visible changes, but tests that assert on persisted state now await `persist.rehydrate()` instead of relying on a synchronous read. The four pre-persist keys (`llmModel` / `providersConfig` / `ttsModel` / `selectedAgentIds`) are deliberately untouched: they are read-only inputs to a migration that never owned them, and `providersConfig` is still read independently by `lib/ai/providers.ts`. Part of #857. * fix(storage): close the async-storage failure modes in the KV persist seam Cross-review found that routing `persist` through an asynchronous store opens failure modes a synchronous `localStorage` never had. Each one below is now covered by a test that fails without its fix. Data loss on a failed or empty hydration. A KV read that threw was indistinguishable from an empty scope, so the store stayed on defaults while an always-on initializer's `set()` persisted those defaults; the next load then read them as a completed migration and deleted the raw original, leaving nothing. Reads now separate "nothing there" from "the backend could not answer": a failed read serves the raw entry read-only, adopts nothing and deletes nothing, and leaves the key unwritable for the session so no un-hydrated state can be persisted over an original. Raw entries are additionally only dropped against a recorded completion marker, so a KV entry of unknown provenance can never authorize deleting the only copy of the user's data. Lost adoption across tabs. A reader whose KV read answered before another tab's adoption write, and whose localStorage read landed after that tab's delete, saw nothing in either place and hydrated defaults. Against a remote backend that interleaving is ordinary rather than rare, so an apparent "nothing anywhere" is now re-checked before it is believed. Out-of-order writes. Two rapid `setItem`s could complete in either order and let the older value win. Reads and writes are serialized per key. Pre-persist migration resurrecting on every load. `migrateFromOldStorage` was guarded by the existence of the raw `settings-storage` key — the very key the migration removes — so once the blob moved into the KVStore it ran on every load and republished whatever the pre-persist keys still held, including API keys the user had since deleted, into first-frame state. It is now the persist layer's last-resort source instead of a synchronous read in the store initializer: its result is adopted like any other migration, so the adopted entry is the marker that stops it. Initial state is plain defaults, and nothing stored reaches observable state before the storage layer has vouched for it. The source keys are still only read, never deleted, since `lib/ai/providers.ts` remains a reader. Also: KV reads and writes are guarded so a corrupt entry surfaces instead of throwing through zustand's silent catch; clearing the cache now clears the migrated stores through their own storage, which a server-backed scope would need and a blanket `localStorage.clear()` would only appear to do; `removeItem` drops the KV copy before the raw one; and the e2e suite gains post-migration steady-state coverage, which every existing spec missed by seeding the pre-cutover key. Part of #857. * fix(storage): scope migration bookkeeping to the device, gate writes on call Second review round. The findings shared one root cause: two of the three sentinels recorded a fact about *this machine's* files but were stored beside the data, in the store's own scope. Under a syncing `account` backend that is a category error with teeth — one device's history would travel to another and authorize deleting files it says nothing about. Sentinels are now always `device`-scoped and carry a reserved `__` prefix, whatever scope the store uses. The worst combination this closes: device A migrates, its marker syncs, device B — which has never run this build — loads and deletes the raw entry that is still the only copy of its own settings. A test drives exactly that sequence against two devices sharing one account scope and nothing else. The pre-persist fallback gains a device sentinel of its own. Its source keys are device-local and deliberately left in place, so "KV and raw are both empty" was not a safe trigger: clearing the account entry, from the cache button or a server-side wipe, let any device re-read those keys and republish credentials the user had rotated. It is now consulted at most once per device, and `removeItem` deliberately does not clear that sentinel — clearing the entry is precisely the case it exists for. The fallback also no longer writes the adoption marker: that marker is a licence to delete a raw persist entry, and the fallback has adopted no such thing. Were it written, a raw entry later restored by an older bundle would be deleted unread. The write gate is now read when `setItem` is called rather than when its queued turn arrives. Evaluated inside the queue, a write issued during hydration waited behind `getItem`, found the gate opened by the very read it raced, and persisted its pre-hydration snapshot on top of the stored value. Refused writes are no longer a dead end: the first one asks the store to rehydrate, since the backend may have recovered, and if that does not work the user is told through a durable notice rather than losing every change on reload. Each refusal is logged, not just the first — a once-per-session warning hides how long a session has been silently unsaved. `removeItem` now propagates backend failures, so clearing the cache cannot report success for a clear that did not happen, and that flow reloads immediately instead of leaving a window in which a background `set()` re-persists the credentials just deleted. Also: the recovery hook is bound after each store is constructed, because naming the store inside its own definition made its type circular and silently widened every selector to `any`; the `serial` comment no longer claims an unreachable rejection handler; and the pre-persist version stamp documents that it skips `migrate`'s unversioned normalization too, which `merge` redoes anyway. Part of #857. * refactor(storage): make the KV persist seam a per-key state machine Three review rounds kept turning up the same defect class in new places: a backend failure read as "there is nothing there", and the result of an operation that nobody looked at. Patching instances of it was not working, because both are possible whenever a call site is free to decide for itself what a failed operation meant. This makes that decision unavailable to call sites. Every backend call now returns an `Outcome` whose payload is a private field and whose only reader demands a `KeyState`. Feeding the machine is not a convention to remember; it is the only way to get the value out. What comes back is the value or `UNAVAILABLE` — deliberately not `null`, so "the backend could not answer" is not spellable as "nothing is there". The machine then decides, in one place, what failure means: any read failure is unavailability, any write failure closes the key, and either raises the health signal and asks for recovery. The scattered closure flags (a settled set, a warned set, a recovery latch) become one explicit phase per key — unhydrated, settled, unavailable, clearing — with the transition table written out in the docstring and a test per row. What that consumes: - A write failing after the key settled used to be ignored; its outcome now closes the key, signals, and the value is remembered for replay. - `removeItem` enters `clearing` synchronously at call time, so a `set()` racing a clear cannot be admitted, queue behind the delete, and write the just-deleted value back. Nothing is replayed across a clear. - The pre-persist sentinel is recorded by *every* settling path, not only the one that consults the old keys. A device that had only ever loaded an already-populated account entry recorded nothing, and so stayed eligible to republish its own stale credentials the moment that entry was cleared — which the docstring already claimed it would not. - Two places treated a failed read as an answer: the sentinel read that decides whether the fallback runs, and the confirming re-read before adopting. Both now stop. - An adoption whose marker did not land no longer settles, and never deletes the raw entry: no proof, no licence. - Refused and failed writes are remembered and replayed once storage recovers, but only when the store was holding real data at the time — replaying a snapshot taken over defaults would overwrite the stored value, which is the failure this seam exists to prevent. When a snapshot cannot be replayed the user is told, because rehydration is about to discard those edits. - Health notices publish on a later task, so a recovery that lands moments later cancels a warning nobody should have seen, and a subscriber mounting after the failure is caught up once its toast host exists. `StorageHealthNotice` now sits after the `Toaster` for the same reason. Also: adoption re-reads before committing so a concurrent write is not rolled back; a shadowed raw entry is only deleted if it parses, matching the promise never to delete bytes we could not read; the unavailable copy points at the control that resolves it rather than suggesting a reload that cannot; and the version pin, the pre-persist retention rationale, and the cache-clear comment say what is actually true. Part of #857. * fix(storage): separate the contexts settle() was conflating Convergence round on the state machine. The findings share a root: settle() treated every read that concluded as licence to open the key, without asking what else was true at the time. - A read finishing while a clear is outstanding no longer settles. Under a remote backend a recovery rehydrate in flight during a clear is ordinary, and opening the gate there admits a write that queues behind the pending delete and restores the credentials just removed. Only finishClear opens a cleared key. - The raw entry served read-only after a failed read no longer counts as the authoritative value. A shadowed raw entry is by construction the older copy — long-term coexistence with a newer KV value is a designed state — so an edit taken against it must never be replayed, or one transient read failure on one device rolls the whole account back. - A write that lands supersedes an earlier failed one. Both were admitted before the key closed, so the queue ordered them; replaying the older snapshot on recovery would undo the newer value. - Unreachable browser storage is a failure, not an empty store. Privacy modes make localStorage throw; treating that as "nothing stored" hydrated defaults, refused every write and said nothing until the user reloaded and found the work gone. SSR stays a silent no-op. - The recovery latch re-arms when storage answers, so the three recovery transitions are not empty promises from the second failure onward. - A refusal carries the phase it happened in. Cold-start timing — an always-on initializer writing before the first read resolves, on a backend that never failed — is a log line, not a sticky red toast; only a refusal under `unavailable` reports lost changes. The health channel stops delivering stale catch-ups: its timers are cancellable and re-read the current state when they fire, so a recovery cannot be overtaken by a warning with nothing left to dismiss it. And `changes-lost` is now a fait accompli rather than a condition — a later recovery has nothing to retract and leaves it standing. Also: a throwing `localStorage` is unavailability rather than absence wherever the raw entry is read; clearing records the pre-persist sentinel like every other settling path, making that docstring true; the raw entry is dropped before the marker so a failed marker clear leaves something inert; and the pre-persist fallback and its four source reads can no longer throw into zustand's silent catch. Part of #857. * fix(storage): hold a replay until it lands, and prove a raw copy before deleting it Three failure-path corners. A replay was retired the moment its write was attempted. Readable but not writable is a real state — a quota-exhausted backend answers reads perfectly — so a recovering read could succeed, the replay write fail, and the snapshot be gone with nothing owed: the next rehydrate served the older stored value over the user's edit and said nothing. The snapshot is now held until the write is durable; a failed replay goes back to being a refused write, so the next recovery tries again and a permanent failure is eventually reported. The adoption marker was treated as proof that a raw entry is the migration's leftover. It proves only that a migration happened on this device once. A rolled-back deployment, or a tab still running the old bundle, writes the raw key again with settings the migration never saw, and the next load on the new bundle deleted them unread. A raw entry is now dropped only when its content equals the value shadowing it — anything else is data, and keeping a copy beats deleting an original here as everywhere else in this file. `changes-lost` was being cancelled by a later recovery. The two statuses answer different questions: "storage is down" describes the world now and stops being true when storage works, while "your edits were lost" describes something that already happened and no amount of recovery undoes. They are now tracked on independent lines per key and rendered as two toasts with distinct ids, so retracting the fault cannot take the acknowledgement with it; the loss notice goes away only when the user closes it. Part of #857. * fix(storage): bound recovery, keep a live store from walking backwards Three interactions the state machine did not have an answer for. Recovery had no budget. A backend that reads but cannot write — a quota-exhausted one does exactly that — turned it into a treadmill: the recovering read succeeds, the key settles, the replay write fails, and that failure asks for recovery again, forever. Attempts now follow a backoff schedule whose length is the cap, and the budget is re-armed only on real progress: a settle that left nothing owed. A settle reached *by* a recovery with a replay still queued is not progress, and counting it as such is precisely what made every lap look like a fresh start. When the budget runs out the key is left read-only and the loss is reported, rather than retrying where nobody can see it failing. A failed read could walk a live store backwards. Once a session has loaded the authoritative value, a transient read failure was still falling back to the shadowed raw copy — which is legitimately older, and legitimately long-lived now that a differing raw entry is kept rather than deleted — and handing it to zustand overwrote a live store with stale data. Worse, the store had already latched "holding real data", so edits made on top of that stale value counted as replayable. A failed read after an authoritative one now returns nothing, which is exactly "keep what you have", and the raw fallback stays where it belongs: the case where nothing authoritative has been seen at all. Dismissing a lost-changes toast only closed the UI. The key stayed in the lost set, so the same store losing changes a second time was swallowed as a duplicate and any later subscriber resurrected a notice the user had already dealt with. The close button now clears the latch, so a second loss is reported and an acknowledged one stays gone. Part of #857. * fix(storage): close the last two provisional and clear-ordering gaps An abandoned migration was handing back an authoritative value. When the confirming re-read fails, or the adoption write does, `adopt` returns the raw envelope unadopted — the same footing as the raw copy served after any other failed read — but it was returned bare, so the store latched "holding real data" and an edit taken on top of it counted as replayable. Another device publishing to the account scope in that window would have been rolled back. `adopt` now reports provisionality like every other read path, and the shape is a named `LoadedValue` so the next path cannot forget. Clearing swallowed a failed raw-entry deletion. The account entry was already gone by then and the marker was cleared afterwards, so the operation reported success while leaving an empty scope beside a live raw entry — precisely the shape the next hydration re-adopts, handing back the settings the user had just deleted. The raw entry now goes first and its failure propagates, which also fixes the ordering: failing before anything is removed leaves the KV value authoritative and the clear simply unperformed, which the caller can retry. Adoption keeps the tolerant variant, where a leftover raw entry is a duplicate rather than a loss. Part of #857. * fix(storage): make the shadowed-raw delete atomic against a concurrent tab cosarah reproduced a data loss: both raw-entry deletes verified the value, then awaited (reading or writing the adoption marker), then removed the key by name. A tab running an old bundle — which takes no Web Lock, so a lock could not have helped — overwrites the raw key with newer settings during that await, and the unconditional `removeItem` then deletes the newer value. The raw side is synchronous `localStorage`, so the fix needs no lock. Both deletes now finish every await first, then run a single synchronous compare-and-delete: re-read the raw bytes, and remove only if they still equal the exact string that was verified. A competing tab is a separate context whose writes surface here only at event-loop boundaries, so its write lands wholly before the re-read (seen — bytes differ — kept) or after the remove, never in between. A changed value is newer data, not a migration leftover, and is kept. Both sites cosarah named are covered: the shadowed-cleanup path (`dropShadowedLegacy`) and the adoption cleanup in `adopt`. `readLegacyEntry` now carries the exact bytes alongside the parse so the re-check is byte-exact rather than comparing a re-serialized value. The clear path's delete is unchanged: it is a deliberate full delete, not a conditional leftover cleanup. Part of #857. * fix(storage): stop deleting the raw entry — adopt by copy (Plan B) Maintainer decision after cross-vendor + red-team review: `localStorage` has no cross-process atomic compare-and-delete (the storage mutex was dropped from the spec), so a tab on an old bundle can insert a write between our re-read and our removeItem in a genuinely parallel process. The synchronous compare-and-delete narrowed that window but could not close it — as its own comment admitted. Since the raw-cleanup cannot be done safely on localStorage while an old bundle may be live, this stops doing it. Adoption is now a copy: the value is written to KV and the raw entry is left in place. The delete race is therefore impossible — there is no delete. KV stays authoritative (once it holds the value, reads take the KV path and never consume the raw one), so the shadow duplicate is inert, left for a wholesale raw-cleanup at migration exit. Removed with the two deletes: - `dropShadowedLegacy` (its only job was the gated delete) and its three call sites. - `deleteLegacyIfUnchanged` and the `{raw,value}` shape of the legacy read (`readLegacyEntry` → back to `readLegacy` returning the value). - The adoption marker (`__persist-adopted:`), whose only reader was that delete gate. It did not prevent re-adoption — the presence of the KV value does — so removing it breaks no invariant. `dropRaw` / `expectedRaw` on `adopt` and the marker-clear step in the clear path go with it. The clear path (user deletes all data) still deletes the raw entry strictly — that is a deliberate full delete, not a migration cleanup. Tests: the delete/marker cases become "the raw shadow always survives, KV is authoritative"; a two-device and an old-bundle-concurrent-write case guard it, and re-adding a delete on either adoption path turns them red. Part of #857. * fix(storage): add a monotonic migrated marker to stop shadow resurrection Cross-vendor + red-team review independently caught a P1 the previous "adopt-as-copy" revision introduced. Because the raw entry is now kept and writes only ever reach KV, the raw copy becomes a frozen pre-cutover value. If the account KV is later emptied by something other than a local clear — a server-side wipe, an account switch, KV eviction — the next load saw `stored === null` beside a live raw entry and re-adopted that stale copy, refilling the wiped account with ancient settings/credentials and syncing them to every device. Deterministic, not a race; it would fire once a server-backed account backend exists. Root cause: "keep the raw entry" inherently requires a record that it has already been migrated, so an empty-KV load can tell "first migration" from "external wipe". Removing the old delete-authorization marker in the copy revision dropped that record — Plan B was only half done. Adds a monotonic, device-scoped `__migrated:<name>` marker: - Read on every load. `stored === null` + raw present + marker set ⇒ respect the wipe, do not adopt (fall through to the confirm re-read, so a concurrent repopulation is still served). Only a genuine first migration (no marker, raw present) adopts. - Written best-effort after a successful KV write. A failed write degrades to the KV value's own presence check, so resurrection needs both a failed marker write and a later external wipe. - Read failure fails closed (do not adopt), never "assume not migrated". - Monotonic: only ever set true, cleared only by a deliberate full clear — no compare-and-delete race, which is why it is safe where deleting the raw entry was not. Live state with a real reader, not the delete-authorization marker removed earlier. The clear path clears the marker too (back to a genuine first-load state); best-effort, since with the raw entry already gone a leftover marker is inert. Tests: resurrection guard (external wipe does not re-adopt; concurrent post-wipe write is served; first migration records the marker; marker read fails closed) + the clear-path marker clearing. Mutation-checked: removing the gate resurrects the shadow, dropping the record leaves it unguarded. P3: the inert `__persist-adopted:` test rot is retargeted to the real `__migrated:` path. Part of #857. * refactor(storage): remove automatic legacy migration — pure KV wiring Maintainer decision (terminal): after repeated review rounds kept finding new P1s in the legacy-adoption path — its state space is too large to keep correct — the PR is narrowed from "migrate + wire" to "wire only". Settings and user-profile persist through the KVStore; pre-cutover data is not migrated; an upgrading user reconfigures once (the data is reconstructable and the cost is one-time). Removed the entire legacy-adoption subsystem from the adapter: - `adopt()`, `readLegacy()`, the `legacyStorage` dep, and the raw-fallback read on a failed KV read. - The `__migrated:` marker and its resurrection guard, the confirm-re-read, and the `provisional`/`LoadedValue` machinery. - The `prePersistFallback` dep, `recordPrePersistHandled`, the `__pre-persist-consulted:` sentinel, and `readSentinel`/`writeSentinel`. - In settings.ts, `readPrePersistSettings` (the llmModel/providersConfig reader wired as that fallback) — reading legacy keys is migration, which this PR no longer does. `lib/ai/providers.ts` still reads `providersConfig` independently, so those source keys are left in place. The adapter now reads and writes the KV scope and touches no legacy key: an empty scope hydrates the store to defaults (ordinary zustand). `load()` is a single KV read; a failed read stays unsettled so defaults are never persisted over an unread value; the clear path removes only the KV entry. Kept, because they are async-storage *steady state* a remote KV backend needs, not migration scaffolding: the KeyState machine (hydration gate, write gate, recovery, replay), the Outcome discipline, per-key serialization, and the persist-health channel. Added `purgeLegacyPersistKey`: a best-effort, fire-and-forget removal of the old raw persist blob, called once by each store. Never read, so a leftover is pure garbage — and the old settings blob held plaintext provider API keys, so clearing it is a small security win. Net: the adapter drops ~460 lines. Tests are rewritten to steady-state KV coverage plus an explicit "old raw key is ignored, not migrated" proof at the store level; the e2e seed fixtures move from the raw key to the KV scope key. Part of #857. * docs(storage): align README and comments with the no-migration behavior Follow-up requested on the approved PR: the migration/adoption logic was removed, but a few comments and one README roadmap line still referred to those deleted concepts. Documentation only — no behaviour change. - README roadmap: the settings/user-profile line said "adopting their pre-cutover raw keys"; it now states there is no automatic migration (legacy keys ignored and best-effort purged, users reconfigure once). - kv-persist.ts: the per-key serial chain comment no longer claims to keep "adoption" from interleaving; it serializes a key's read/write/clear. - settings.ts: `SETTINGS_PERSIST_VERSION` no longer documents a "pre-persist fallback" that no longer exists. - Test/fixture headers that described seeding "the pre-cutover raw key" for "one-time adoption" now describe seeding the KV `account` scope, with no legacy-key migration. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * chore(storage): bump version to 0.1.1 for README doc alignment The README is a publishable input, and the no-migration doc alignment changed it, so the package version must increase. Patch bump: documentation only, no code or API change. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com> Co-authored-by: 杨慎 <117187635+cosarah@users.noreply.github.com> Co-authored-by: wyuc <zdq1204@gmail.com> |
||
|
sophietao20-star
|
0cf2a33041 |
docs: refresh and synchronize documentation (#996)
Refresh the Chinese documentation against current functionality, synchronize all localized guides, document deployment and configuration behavior, and align the PPTX import feature flag implementation. |
||
|
LING DUAN
|
61a7154f07 |
feat(pi): add evidence-aware Director context runtime (#971)
* feat(pi): add evidence-aware Director context runtime * fix(web-search): require completed native search * fix(orchestration): preserve whiteboard across agent handoffs * fix(pi): remove wrap-up turn bypass |
||
|
|
b4e3145c90 |
feat(roster): single-source the generated agent roster on the stage document (#994)
* feat(roster): single-source the generated agent roster on the stage document The generated agent roster used to live in two places at once: the stage document carried generatedAgentConfigs, while a per-stage IndexedDB mirror (db.generatedAgents) was the actual read authority — and the only home of each agent's voice binding. The split caused real losses: voices vanished on device switches, exports, and shares (the mirror never travels), the load path trusted a device-local table over the portable document, and roster edits persisted through a bespoke debounce with no beforeunload flush. Make the stage document the single source of truth: - Contract: add optional voiceConfig/voiceDesign to GeneratedAgentConfig in @openmaic/dsl (VoiceDesign moves into the contract; the app re-exports it). Additive optional fields — not a breaking serialized-shape change, so DSL_VERSION stays put per the version policy. - Writes: remove every mirror writer (registry saveGeneratedAgents, the stage store's debouncedSaveAgents bypass, classroom-load hydration, import direct writes). setStageAgents now only marks the stage dirty for the shared persistence scheduler plus synchronous in-memory registry/selection mirrors; generation and import embed the roster on the stage. - Reads: classroom load hydrates the registry from the loaded stage's generatedAgentConfigs; export reads the stage roster directly. - Lazy migration: the mirror is retained read-only. On load, a roster (or voice fields) missing from the document is backfilled from the mirror, committed onto the in-memory stage, and persisted by the next flush — idempotent, and gated on the current stage id after every await so a classroom switch cannot leak a stale roster into the registry. - Deletion: deleteStageData discards the deleted stage's pending persistence work so a queued flush cannot resurrect the document. - Round trip: classroom ZIP manifests now carry the voice fields, so export/import preserves agent voices. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(roster): harden deletion, migration, selection, and ingress paths Review hardening on top of the roster single-source change: - Deletion vs in-flight persistence: deleting a stage now registers a session-scoped tombstone (lib/utils/deleted-stages.ts) before the cascade starts. Every persistence landing point checks it — the scheduled flush and the departing-stage retry drop their snapshots, and the aggregate/incremental saves re-check under the document lock, including the incremental path's full-save fallback that would otherwise rebuild a deleted document from the in-memory snapshot. A deletion that fails before the document is removed lifts the tombstone so the still-existing stage keeps persisting. - Lazy migration termination: a fruitless legacy-mirror probe (nothing to merge — e.g. server-generated rosters, which are voiceless by design) is remembered per stage for the session instead of re-querying the mirror on every load forever. Successful merges are not memoized so a failed flush is retried on the next load. Docstrings now state the real lifecycle. - Selection provenance: setStageAgents no longer overwrites a user-set agent selection with the full roster. Stage-derived selections still track the roster; a user-set auto selection is only narrowed to surviving agents (newcomers are not auto-selected, the user-set flag is untouched), and a user-set preset selection is left alone — matching restoreAgentSelection. - Mirror hygiene: deleteStageData clears the deleted stage's rows from the legacy roster mirror (best-effort), and the mirror table's docs now describe its actual access pattern (migration reads + deletion hygiene). - Registry persistence: the agent registry's persist snapshot partializes generated agents out of localStorage, making the stage document the roster's only durable home; the rehydration merge filter remains as defense in depth. - Ingress validation: imported manifest voice fields are structurally validated (malformed bindings dropped per field, the agent survives), and applyGeneratedAgentsToRegistry validates providerId against the known TTS provider registry instead of casting, treating unknown providers as "no bound voice". - Contract: drop the speculative AgentVoiceConfig.modelId — an audit of every roster producer, current and historical, found none that ever emitted it. Contract fields are added once a producer exists, not before. New tests cover the deletion tombstone at both the store and storage layers (including the missing-destination full-save pin and the lock-race re-check), the fruitless-probe memo, the selection-provenance matrix, persistence exclusion of generated agents, and voice-field sanitization on import. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(roster): close tombstone lifecycle, probe-memo, and selection-intent gaps Second hardening pass over the deletion/migration/selection fixes: - Tombstone lifecycle: a deleted classroom id can legitimately come back (deletion is client-side only). Explicit (re)creation points now lift the tombstone — server-copy restore in applyClassroomStageAndScenes and backup restore in importDatabase — so edits after a same-session restore persist again, while in-flight flushes stay fenced. - Failed deletion now restores the dirt it discarded: deleteStageData snapshots the pending map plus the in-flight round before tombstoning, and re-marks it when the delete fails with the document still present. - Legacy-mirror probe memo now distinguishes a FAILED read (null, retried on the next load) from a confirmed-empty mirror (memoized), so a transient IndexedDB error cannot suppress migration for the whole session. - Incremental save tail (currentScene KV row + chat sessions) is tombstone- fenced like the aggregate path, and both paths re-check the tombstone immediately before every write, with comments scoped honestly for the lock-free (no Web Locks) best-effort LWW fallback. - Roster-edit selection mirror: an empty intersection now falls back to the full roster and clears the user-set flag (matching restoreAgentSelection's length > 0 gate), and a user-set selection that equals the pre-edit full roster keeps tracking the roster wholesale so newcomers are not excluded forever by the AgentBar auto-toggle snapshot. - TTS/ASR provider lookups use Object.hasOwn instead of `in`/bare indexing, so prototype-chain keys ('toString', 'constructor', ...) no longer pass the provider whitelist or resolve to Object.prototype members. - Documented the per-tab limit of the tombstone set. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(roster): replace the boolean deletion tombstone with a deletion-generation model A boolean tombstone cannot express the deletion lifecycle once a deleted id legitimately comes back (server-copy restore on revisit, backup import): after the restore lifts the flag, a pre-delete flush still in flight is indistinguishable from a post-restore edit and can overwrite the restored document with pre-delete content. Two rounds of point fixes kept leaking variants of this race, so this restructures the fence instead. deleted-stages.ts now keeps a per-stage monotonic deletion epoch plus a deleted flag. markStageDeleted bumps the epoch; unmarkStageDeleted clears only the flag and never rewinds the epoch. Every persistence path captures the epoch at the moment it captures the data it will write (flush-round snapshot, departing-stage snapshot, aggregate save entry) and re-checks "captured epoch is current AND not deleted" immediately before each individual write. Invariants: - A write captured before a deletion can never land after it, even when a same-id restore has lifted the deleted flag (the captured epoch is permanently stale). - A write capturing data after a restore observes the current epoch and persists normally. - A failed deletion lifts only the flag; the restored pending changes are re-queued as descriptors, so their flush re-captures current store state under the current epoch and is neither dropped nor a stale replay. Also fixed in the same restructuring: - Incremental tail writes (currentScene, chats) each re-check independently; a delete landing while the first tail write is awaiting now fences the chat write instead of riding the earlier check. - Deleting a classroom evicts it from the warm in-memory store, and loadFromStorage treats a warm-but-deleted stage as not loaded (discarding the ghost), so navigating Back to a deleted classroom reaches the server-restore path instead of rendering an editable ghost whose every edit is silently dropped. - importDatabase records each document's pre-import deletion state and reinstates it when a failed import rolls the document back, so an outstanding flush cannot recreate the rolled-back document. - Docstrings on the failed-deletion restore path now state that only dirt captured before the deletion is restored; edits refused during the deletion window are not part of the snapshot. New tests cover the epoch invariants at the storage layer (pre-delete rounds dropped across restores on the incremental, aggregate, and tail paths; post-restore captures landing; strictly increasing epochs), the scheduler layer (departing-retry delete+restore straddle; restored dirt flushing under the current epoch), the warm-ghost restore driven through the real loadFromStorage, and the import lift/rollback pair. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(deletion): report fenced drops, gate ghost eviction on cascade settlement Edge hardening around the deletion-epoch fence; the epoch model itself is unchanged. - An epoch-stale drop is now reported, not disguised: saveStageData / saveStageDataIncremental return a distinct 'stale-dropped' status (entry check, every in-mutation re-check, and both tail-write re-checks) instead of a shape identical to success. saveToStorage reacts by skipping all success bookkeeping - no chatSnapshot rebind to a snapshot that never landed, no pending-dirt clearing - and returns false, keeping its "true means verified write" contract honest. persistDirtySnapshot keeps its existing (correct) semantics explicitly: a drop has nothing to retry. - Warm-ghost eviction is settlement-gated: the deletion cascade records an in-flight bit (begun by deleteStageData, settled in its finally), and the deleted-warm branch of loadFromStorage keeps the warm state while the cascade is unsettled - a delete can still fail before removing the document, in which case the warm state plus the restored pending dirt is the only copy of the pre-delete edits. Once settled, the ghost discard and server-restore path behave as before. - Read-side landing re-check: loadFromStorage re-checks isStageDeleted after hydration, immediately before set(), mirroring the write-side "re-check immediately before landing" discipline - a lock-free mid-cascade read can no longer re-materialize a ghost classroom. - clearStoreForDeletedStage re-checks isStageDeleted at eviction time, so a same-id restore completing inside the cascade-tail window is never wiped. - capturedEpoch is now a required parameter on both storage entry points: the default (call-time capture) silently reopened the capture-point/ validation-point split for future callers; the type system now enforces the pairing. All production callers already passed it explicitly. New tests cover the dropped-status contract, the cascade in-flight lifecycle (including both failure shapes), keep-warm during an unsettled delete with end-to-end pending restore after a failed delete, settlement- then-eviction on success, the eviction guard against a tail-window restore, and the read-side hydration re-check. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(deletion): settle the deleted-warm load by awaiting the cascade outcome Hardening around the deletion settlement gate introduced previously; the epoch model is unchanged. - The deleted-warm branch of loadFromStorage no longer completes the load while the deletion cascade is undecided (which exposed a classroom whose edits were refused without being restorable, and left a successful deletion's eviction with no follow-up reload). It now parks on a per-stage settlement promise (stageDeletionSettled) and branches on the outcome: a failed delete keeps the warm, restored classroom; a successful delete falls through to a full cold load so the server-restore path recovers the route; a navigation that moved on during the park is stopped by the load token guard. The eviction (clearStoreForDeletedStage) deliberately stops claiming a new load token so the parked load stays current - ghost re-materialization is fenced by the read-side deleted re-checks instead. No deadlock: a parked load holds no document lock, and the cascade never waits on a load. - 'stale-dropped' now propagates through the debounced flush path: persistDirtySnapshot returns the sentinel instead of an empty failure set, and startFlushRound skips the chatSnapshot rebind on it (mirroring saveToStorage) - a falsely rebound baseline would let later chat saves no-op-skip chats that never landed and corrupt cross-tab conflict detection. Pending-map clearing keeps its empty-failure semantics (restored descriptors mint fresh revisions). - The cascade in-flight state is a counter (begin++/settle--) instead of a boolean, and deleteStageData is single-flight per stage (a concurrent second call joins the first cascade), so overlapping deletions can never expose an undecided cascade as settled. - The StaleDroppedSave docstring no longer overpromises: a tail saveCurrentScene can land after the cascade's clearCurrentScene; the orphaned cursor row is ignored by the load path rather than removed by the cascade. Net +8 tests: overlapping-cascade counting, settlement-promise resolution (success and failure), concurrent double-delete single-flight join, the parked mid-cascade Back in both outcomes plus the navigation-away guard, and the fenced flush round keeping an honest chat baseline end-to-end. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(deletion): re-verify store identity on keep-warm resume, re-run joined deletes after a mid-cascade restore Three hardening refinements on the deletion/persistence fencing: - loadFromStorage's parked keep-warm branch now mirrors the success path's store-identity guard: if a tokenless writer (e.g. a database import) replaced the stage during the park, a failed deletion falls through to the cold load instead of reporting another classroom as the kept warm state. The ghost-discard block is correspondingly gated on the deleted flag so the fall-through does not mislabel a surviving document as a ghost. - deleteStageData: a joined call carries its own deletion intent. When the first cascade fulfills but a same-id restore lifted the deleted flag in the meantime, the joined caller now runs one fresh cascade against the restored document instead of reporting the pre-restore outcome. Exactly one re-check per call (no recursion); a rejected first cascade still propagates unchanged to every joined caller. - Narrowed the failed-delete restore-ownership comments: the restore covers only the pending map plus an in-flight flush round's dirt; a departing-stage snapshot is outside that capture and is fenced and dropped by design. Comment-only, no behavior change. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(roster): gate deleted-warm handling on stage identity, keep an explicit empty roster authoritative Final-audit fixes, three items: - loadFromStorage keyed its deleted-warm settlement handling (park / keep-warm / ghost discard) behind the scenes.length > 0 shortcut, so a zero-scene warm ghost of a deleted classroom skipped it entirely: the cold load found no document, the ghost stayed in the store, and the classroom loader's server-fallback gate (!getCurrentStage()) never ran — the silently-uneditable trap this branch exists to close, in its scene-less form. Deletion handling now keys on stage identity alone, before any scene-count shortcut; the non-deleted warm skip still requires scenes. - rosterNeedsLegacyFallback collapsed "roster field absent" (document predates roster persistence) with "explicitly persisted empty roster". A future writer persisting [] on a device whose read-only mirror still held stale rows would have its emptied roster resurrected on every load. The full-lift branch now triggers only when the field is absent (undefined); an explicit [] is authoritative and never probes the mirror. No current writer produces [], but the read side no longer depends on that invariant. - The GeneratedAgentConfig docstring claimed the voice fields are not a breaking change; that only holds for this codebase's tolerant structural validators. The generated stage.schema.json sets additionalProperties: false, so consumers pinning an older published schema artifact reject documents carrying the new fields. The docstring now states both sides honestly. Tests: a zero-scene deleted warm ghost is restored through the real loadFromStorage (fails pre-fix), an explicit empty roster does not resurrect stale mirror rows (fails pre-fix), and the rosterNeedsLegacyFallback unit matrix covers undefined vs []. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(deletion): re-queue aggregate persistence when a failed deletion is recovered Direct aggregate saves (saveToStorage: generation completion, server-restore hydration) run outside the pending map and the flush round, so the deletion path's pre-delete dirt snapshot cannot describe them. When a deletion bumped the epoch while such a save was in flight, the save was correctly fenced ('stale-dropped') — but if the deletion then failed before removing the document, the recovery only re-queued the snapshotted scheduler descriptors. The fenced aggregate-only content stayed memory-only with no retry anywhere, and a later reload lost it. The failed-deletion restore now merges a full re-mark of the aggregate (structure / stage / outline / currentScene / chats plus a descriptor for every current scene) into the recovery set, still guarded on the store holding the stage. Because the flush recaptures the CURRENT store state under the CURRENT epoch, this necessarily carries whatever the fenced aggregate save held — and, for the same reason, edits refused by the scheduler during the deletion window, which previously stayed memory-only as well; the scope docstrings now say so instead of declaring them excluded. New integration coverage drives the real store scheduler and the real storage layer end to end: a genuinely fenced in-flight saveToStorage (generation completion) plus a pre-removal deletion failure ends with the completion flag durable on the next flush; a deletion-window scene edit survives the same way; and a successful deletion re-marks nothing (control). Both recovery tests fail against the previous restore logic. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com> Co-authored-by: 杨慎 <117187635+cosarah@users.noreply.github.com> |
||
|
wyucandCodex
|
e39c64cd9e |
feat(persistence): one-command server-backed stack (compose profile + embedded API + docs) (#982)
* feat(persistence): one-command server-backed stack — embedded API + compose postgres profile docker compose --profile server-persistence up --build runs exactly two containers: the app (persistence HTTP server embedded at /api/persistence, enabled only when DATABASE_URL is set — unset keeps today's browser-only behavior byte-for-byte) and PostgreSQL 16. Client bootstrap configures both storage seams against the same-origin route when NEXT_PUBLIC_PERSISTENCE=1, riding the existing lazy migrations so prior browser data reaches the server per-course on first open. Dev auth (bearer token + client-asserted learner header) lives in one overridable module and is loudly documented as development-grade. Verified end-to-end against the real stack: document PUT/GET and runtime session create land in Postgres; bad tokens 401. Co-authored-by: Codex <codex@openai.com> * fix(persistence): review round — structural bootstrap ordering, retryable init, honest failure surfaces The bootstrap side-effect import moves into the two storage seam entry modules, so any module graph that can resolve a store necessarily evaluates it first — no client-entry ordering luck, and the two configure calls are all-or-nothing. A failed server init clears its cached promise (next request retries instead of permanent 500s). listStages rethrows backend failures and the dashboard surfaces a persistence-unavailable error instead of rendering an empty list that reads as data loss. Docs state the build-time nature of NEXT_PUBLIC_PERSISTENCE, the real token boundary (any visitor can extract it and impersonate any learner — localhost/trusted-network only), and the correct postgres password rotation. Dev-token compare is timing-safe; the handler singleton lives on globalThis so HMR cannot leak pools. Co-authored-by: Codex <codex@openai.com> * fix(persistence): review response — no-confidentiality wording + adapter round-trip coverage The token docs now state plainly that the NEXT_PUBLIC token is compiled into the public bundle and provides no confidentiality or user isolation (its only role is keeping unrelated scanners off a trusted network). The Fetch/Node adapter — the route's most bug-prone code — gains a round-trip integration test: PUT body streaming, 201 with multi-value headers, 204 empty body, and encoded path segments reaching the handler un-decoded. --------- Co-authored-by: Codex <codex@openai.com> |
||
|
杨慎andClaude Opus 4.8
|
84b1907255 |
feat(video-export): service-backed MP4 render + in-app one-click export (#866) (#937)
* feat(video-export): service-backed MP4 render + in-app one-click export (#866)
Adds the last mile of classroom video export: turning the self-contained
Hyperframes project ZIP (#865) into an MP4 via an isolated render service,
one-click in-app.
- render-service/: standalone Node 22 + Chromium + FFmpeg container wrapping
@hyperframes/producer's library API. Async job model (POST /render -> 202
jobId, GET poll, GET download, DELETE cancel). Swappable JobStore /
ArtifactStore seams (in-memory + local-disk now; Redis/S3 + presigned-302
download later) so it scales horizontally without changing the HTTP contract.
Concurrency + per-user guards are config knobs.
- App integration: thin Next proxy routes under app/api/export-video/* (forward
only, no rendering) + capability probe. use-render-video.ts uploads the ZIP,
polls via runPolledTask, downloads the MP4; shared buildExportZip prefix with
the existing ZIP path. Export menu gains resolution/fps/quality selectors and
a progress bar; degrades to ZIP download when RENDER_SERVICE_URL is unset.
- docker-compose: render-service under an opt-in "video-export" profile.
- Entry is main.ts (not server.ts): the producer auto-starts its own server on
:9847 when the process entry path ends with /src/server.ts.
Verified end-to-end in the container: rendered a real 640s (10.7 min) classroom
ZIP to a valid H.264 720p + AAC MP4 (duration matches source) in ~9.6 min
(~0.9x realtime, 4-worker frame capture). Degrade path, queued-cancel + cleanup,
and per-user 429 guard all exercised. pnpm check / lint / tsc / i18n pass.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* feat(video-export): global render progress store, percent+ETA UI, ring on export button (#866)
Addresses two UX issues found while driving the in-app MP4 export:
1. Progress display was raw and unfriendly (showed producer's English stage
strings like "Capturing frame 5130/19220") and had no time estimate. Now the
menu shows only "<percent>% · about <remaining> left". ETA is computed from a
recent-speed estimate (percent-per-ms over the last sample), EMA-smoothed —
which tracks the render's non-uniform pace (prep -> frame capture with a
4->1 worker drop -> encode) far better than a whole-run average, and never
shows a stale/rising ETA.
2. Switching scenes mid-render unmounted the export menu and lost the progress
(and reset the local "already rendering" ref, allowing a duplicate submit).
The whole render lifecycle now lives in a global store
(lib/store/video-render.ts), so progress survives menu close / scene switch
and duplicate submits are guarded by status. A persistent CircularProgress
ring on the export button shows live progress whether or not the menu is open.
Also fixes the progress scale: the producer reports progress as 0..100, but our
HTTP contract (and success path) is 0..1 — the service now normalizes it, so the
client no longer showed "2000%".
- lib/store/video-render.ts: new global store owning submit->poll->download,
recent-speed ETA, duplicate-submit guard.
- lib/video-export-app/use-render-video.ts: thin facade over the store.
- components/ui/circular-progress.tsx: lightweight SVG progress ring.
- components/stage/{header-controls,video-export-menu}.tsx: ring on the export
button; menu shows percent + ETA, subscribes to the store.
- render-service/src/render-manager.ts: normalize producer progress 0..100 -> 0..1.
- i18n: percent/ETA strings across all 8 locales (drops the stage-based string).
- render-service/package-lock.json: complete integrity hashes (reproducible npm ci).
Verified: ETA logic checked against the real segmented render curve (worker drop
raises ETA, encode speedup drives it to ~0); progress scale fix confirmed live
against the container (0.2 -> 20%). tsc / lint / prettier / i18n pass.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* fix(video-export): persist render options in the store, not the menu component
Selecting 720p/24fps/draft, switching scenes, and reopening the export menu
showed the defaults again (1080p/30/standard). The selections lived in the
VideoExportMenu component's local state, which reset when the menu unmounted on
a scene switch — the running render still used the chosen options, but the UI
misrepresented them.
Move resolution/fps/quality into the global video-render store (with a
setOptions action). The menu now reads/writes the store, so selections survive
menu close / scene switch, and while a render runs the selectors reflect the
options that render is actually using. startRender() reads options from the
store instead of taking them as an argument.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* fix(video-export): deployment correctness + resource/isolation controls (PR #937 review)
Addresses the blocking findings from wyuc's review. Output fidelity was fine;
these harden deployment and production resource/isolation boundaries.
#1 Compose advertised MP4 but couldn't render in prod:
- Capability now probes the service's /health (checkRenderServiceHealth), so a
configured-but-absent service reports disabled and the UI degrades to ZIP
instead of 502-ing.
- RENDER_SERVICE_URL is operator-supplied trusted config, so the proxy no longer
runs it through the SSRF guard — the one-command `docker compose --profile
video-export up` now works without globally weakening SSRF via
ALLOW_LOCAL_NETWORKS. resolveRenderServiceUrl() is now synchronous.
- Client degrades to ZIP on any failed submit (not only 501).
#2 Unbounded upload/queue (ZIP-bomb / DoS):
- unzip.ts bounds the archive via fflate's filter BEFORE decompression: entry
count, per-entry and total expanded size, and compression ratio.
- Proxy rejects oversized uploads (413) by Content-Length before forwarding.
- RenderManager enforces a global queue-depth cap (RENDER_MAX_QUEUE).
- All limits are env-tunable knobs in config.ts.
#3 Per-user guard was ineffective + admission ran after extraction:
- Identity is derived server-side (client IP) and forwarded as x-openmaic-client;
the service ignores any client-supplied userId, and the proxy strips it.
- Admission is split into reserve()/submit()/release(): the slot is reserved
BEFORE extraction, so a rejected caller never triggers a decompression.
Additional risks:
- Per-job wall-clock watchdog (RENDER_JOB_DEADLINE_MS) aborts + fails a hung
render so it can't hold a slot/scratch forever.
- Download proxy bounds only the time-to-headers, not the body stream, so large
MP4s over slow links no longer truncate.
- Client cancels the server job (DELETE) when a started render fails/times out.
- Compose puts render-service on an internal:true network (no host/internet
route), sandboxing the Chromium that runs the uploaded HTML; the export ZIP is
self-contained so no outbound is needed. README documents the standalone caveat.
Not closing #866: the smoke/golden-render CI acceptance criterion remains a
follow-up (see PR description).
Verified in-container: legal render 202; ZIP-bomb (entry-count + compression-
ratio) rejected 400 before any decompression; per-identity guard 429 with a
spoofed multipart userId ignored; reserve-before-extract leaves no scratch dir
on rejection; watchdog aborts an overrunning job and frees the slot. tsc / lint /
prettier / i18n pass; render-service tsc passes.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* fix(video-export): real audio durations + burned-in subtitles (PR #937 review)
Two export-fidelity issues found in wyuc's deeper E2E:
A. Narration was scheduled from estimated durations, cutting audio off mid-
sentence and advancing the timeline early. The scheduler trusted
AudioFileRecord.duration (recorded only since #861), so the many existing
classrooms without it fell back to text-length estimates — measured 4.35s
average / 10.23s max underestimate across 47 clips. timeline-deps now probes
the real duration from each narration blob via an off-document <audio>
(symmetric to the existing video probe), preferring it over the stored
duration, then the estimate only when no audio asset exists. Everything
downstream (narration starts, scene/total duration, subtitle cues) re-derives
from the corrected value in the pure compiler — no compiler change needed.
B. The final MP4 had no subtitles (only H.264+AAC), and the ZIP's SRT/VTT used
the same estimated boundaries. The emitter now renders a burned-in subtitle
overlay: one caption box + a hidden div per cue, revealed/hidden by the paused
GSAP timeline at each cue's start/end (corrected timings from A), so Chromium's
frame capture bakes them in. The producer has no subtitle track of its own, so
burn-in is the v1 approach.
Verified: emitter unit tests + snapshot updated (subtitle overlay + toggle
statements, escaped text, hidden-by-default); 82 video-export tests pass incl.
the determinism red-line proxy. Rendered a synthetic subtitle project through the
container and confirmed by pixel analysis that captions appear only within their
cue window (2429 near-white px in the caption band at t=1.5s vs 0 at t=0.05s).
tsc / lint / prettier / i18n pass.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* fix(video-export): subtitle layout + upload/admission hardening (PR #937 review)
Address the three P1 blockers plus actionable P2s from the
|
||
|
|
bb8d4f30e7 |
feat(chat): add experimental Pi classroom runtime behind a flag (#914)
* feat(chat): add experimental Pi classroom runtime * fix(chat): harden Pi whiteboard action sequencing * style(chat): format session retirement flow * style(playback): format async scene transition * fix(chat): retire actions on session errors * fix(chat): close Pi runtime lifecycle gaps * fix(chat): avoid replaying historical whiteboard actions * fix(chat): prevent Pi structured-output residue from leaking and inflating turns Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com> * fix(chat): expose whiteboard element/line ids and bound code-line context Surface each element's id (and code line ids) in the Pi whiteboard context so later child agents can target wb_delete / wb_edit_code, and cap the rendered code lines with a shared character budget so a code-heavy board cannot grow the prompt without bound. Return only the current turn's whiteboard ledger from the Pi director loop instead of carrying history forward across requests. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com> * fix(chat): bound Pi whiteboard code context Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com> * fix(chat): prioritize newer persisted whiteboard code in budget Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com> * fix(chat): enforce whiteboard boundaries across Pi sessions * feat(chat): refine soft-closing discussion controls * fix(chat): make whiteboard clearing model-directed --------- Co-authored-by: apple <apple@GigideMacBook-Air.local> Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com> Co-authored-by: LING-6150 <duan.lin@northeastern.edu> |
||
|
Yizuki_Ameandwyuc
|
7c2aaafc91 |
fix(ssrf): harden provider redirects and ISATAP detection (#928)
* fix(ssrf): detect private IPv4 in ISATAP addresses * fix(ssrf): reject redirects in provider probes --------- Co-authored-by: wyuc <wang-yc24@mails.tsinghua.edu.cn> |
||
|
|
cc5a6ab152 |
feat(agent): edit_elements — NL → EditIntent vertical (#895) (#896)
* feat(agent): add edit_elements NL → EditIntent vertical (#895) Give the editing surface its first real consumer: natural-language per-element style/geometry edits flow through the same EditIntent vocabulary as pointer gestures, with an all-or-nothing validation gate and one undo entry per instruction. Co-authored-by: Cursor <cursoragent@cursor.com> * fix(agent): harden edit_elements against CR findings (#895) Re-validate intents at apply time (atomic refuse on missing/locked ids), type-check prop values, require an open edit session for undo, clamp line stroke width correctly, enforce group cohesion, nest shape text chrome, and surface refuseReason on the tool card. Co-authored-by: Cursor <cursoragent@cursor.com> * fix(agent): close remaining edit_elements CR residuals (#895) Feed apply outcomes into next-turn history and post-run chat corrections, phrase tool success as proposed-not-applied, and tighten gradient/filters/ enums/per-type value validation plus label-less shape text refusal. Co-authored-by: Cursor <cursoragent@cursor.com> * refactor(agent): derive edit_elements prop validation from the DSL schema (#895) Three review rounds each found a hand-written validator disagreeing with the DSL contract (a gradient pos range pinned at [0,1] against the percent contract, a textType enum matching the DSL in neither direction, per-type ownership gaps like vertical-on-shape and color-on-everything). Replace the class instead of the instances: - The gate now loads the dsl-generated JSON schema (@openmaic/dsl/schema/scene.schema.json) and validates every proposed prop value against the target element type's subschema via a compact schema-subset checker — per-type ownership, enums, and structured shapes (outline/shadow/gradient/filters) come from the single source of truth and stay correct as the DSL evolves. - Explicit policy layers stay on top: the content/identity blocklist (schema would allow what policy forbids), geometry clamps (coords, min sizes, line stroke width), and a numeric-range overlay for what plain number types cannot express — gradient stop pos is a percent in [0,100] (the previous [0,1] check refused canonical gradients and the unit test pinned the wrong contract), fontSize/lineHeight/radius bounds, filter value length, stop-count cap. - Consequences: color is refused on types that do not render it, with prompt guidance steering text color to defaultColor and shape body to fill; fill on charts is now allowed (the schema has it; the inventory already advertised it); textType accepts the full DSL enum. - Apply-time revalidation also refuses element-type drift via a gate-time targetElementTypes map carried in the tool details. - Refusal reasons render as visible card text, not only a hover title. - Test pins added for all of the above plus selection scoping, coord bounds, non-finite rotate, and empty batch/props. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(agent): harden the schema-derived gate — fail closed, assert schema source, unify numeric clamping (#895) - A schema construct the subset checker cannot judge (unresolved $ref, malformed anyOf, unsupported keywords) now refuses instead of accepting — the gate fails closed on the unknown. - A module-load assertion verifies the loaded dsl schema carries every definition the gate maps and a sane TextType enum, so a stale or missing build artifact fails loudly with a rebuild hint instead of validating an outdated contract silently. - Numeric-overshoot convention unified and documented: structural, enum, or ownership mismatches refuse (teaching the model the contract); numeric overshoot on a valid prop clamps (matching gesture behavior) — opacity now clamps like geometry. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(agent): localize edit element refusals (#895) * fix(agent): preserve partial element edit semantics (#895) * fix(agent): prevent stale element edit application (#895) * fix(agent): preserve live canvas edit invariants (#895) * fix(agent): close element style contract gaps (#895) * fix(agent): reject non-persistent element styles (#895) * fix(agent): preserve rigid group geometry (#895) * fix(agent): tolerate decimal group translations (#895) * fix(agent): constrain line height edits (#895) * fix(agent): reject stale or shadowed element edits (#895) * fix(agent): close renderer contract gaps (#895) --------- Co-authored-by: Cursor <cursoragent@cursor.com> Co-authored-by: Claude Fable 5 <noreply@anthropic.com> Co-authored-by: 杨慎 <117187635+cosarah@users.noreply.github.com> |
||
|
Yanpeng Wangandwyuc
|
9670100363 |
feat(extraction): media (audio/video) extraction + AliDocMind provider (#887)
* feat(extraction): add AliDocMind provider + media extraction abstraction Adds AliDocMind (Aliyun Document Mind LLM version) as a new vendor alongside unpdf/MinerU, and introduces the media (audio/video) extraction layer that mirrors the document extraction one. Document side (file: pdf/docx/pptx/xlsx/images): - PDF_PROVIDERS gains an `alidocmind` entry; parseWithAliDocMind() maps layouts[] -> ParsedPdfContent. Flows through the existing document extractor registry, so AliDocMind is selectable anywhere MinerU is. - PDFParserConfig / DocumentExtractorConfig gain accessKeyId/accessKeySecret (AliDocMind uses AK/SK, not a single apiKey); env fallback via ALIDOCMIND_ACCESS_KEY_ID / ALIDOCMIND_ACCESS_KEY_SECRET. Media side (mp4/mp3/wav/... -> MediaArtifact): - New lib/media-parse/ domain mirroring lib/pdf/ (types/constants/providers). parseMedia() maps AliDocMind segments[]/audio_frames/video_frames -> MediaArtifact (transcript + keyframes). - MediaExtractorProvider interface + media registry + extractMedia() entry, symmetric to DocumentExtractorProvider / extractDocument(). - MediaArtifact and the ExtractionResult/Artifact/Error/Job envelope live in lib/document/types.ts (re-exported from @/lib/document). Shared AliDocMind SDK wrapper (lib/pdf/alidocmind-client.ts) handles the submit -> poll -> get flow for both sides via @alicloud/docmind-api20220711. Tests: env-gated smoke test (tests/document/alidocmind.smoke.test.ts) drives a real PDF and a real video through AliDocMind; media-artifact type test; extractor-registry test updated for the new provider. Part of #621 (MAIC ETL). Media extraction is the sibling to the document extraction landed in #741. * feat(extraction): wire AliDocMind AK/SK through settings UI + routes Surfaces AliDocMind in the (post-#837) Document Parsing settings panel as a peer of unpdf/MinerU — one panel, one credential entry, more supported formats. Threads Aliyun AccessKey ID/Secret from the store through the extraction routes. - Store: pdfProvidersConfig + setPDFProviderConfig gain accessKeyId/accessKeySecret; default alidocmind entry added. - pdf-settings.tsx: AliDocMind branch renders AccessKey ID + Secret inputs (secret masked with show/hide) and a Test Connection button; request-URL preview shows the DocMind endpoint. Provider auto-appears in the panel via PDF_PROVIDERS; supported-format badges come from #837's registry (ALIDOCMIND_MIMES added to lib/document/mime.ts). - verify-pdf-provider route: alidocmind branch verifies AK/SK via a lightweight authenticated probe (verifyAliDocMindCredentials) — auth-level errors fail, anything else passes. - extract-document route + generation flow (app/page.tsx, generation-preview): accessKeyId/accessKeySecret carried through session → FormData → config. - i18n: alidocmindAccessKeyId / alidocmindAccessKeySecret in 8 locales. - Icon: reuse /logos/bailian.svg (Aliyun family) instead of a missing asset. Verified end-to-end against the running app: AliDocMind panel renders, Test Connection returns "连接成功" through the real Aliyun API. Part of #886. * feat(extraction): route audio/video uploads through extractMedia() (reuse document path) Media uploads now flow through the same upload picker, /api/extract-document route, and generation pipeline as documents — no separate upload area. Only the extraction differs: media mimes dispatch to extractMedia() -> MediaArtifact, which is flattened to the text shape the generation pipeline already consumes. - mime.ts: register audio/video formats in DOCUMENT_FORMATS (accept string, extension map, badges resolve for them); add MEDIA_PROVIDER_SUPPORTED_MIME_TYPES + SUPPORTED_MEDIA_MIME_TYPES, kept separate from PROVIDER_SUPPORTED_MIME_TYPES so the document drift-guard stays document-only. mimesForProviders() folds in a provider's media mimes, so the existing upload helpers (getAcceptStringForProviders / isMimeSupportedByProviders / format badges) cover media automatically when the provider supports it. - extract-document route: media mimes dispatch to extractMedia(); MediaArtifact flattened to timestamped text (synopsis + transcript + keyframes). - Tests: 6 media cases in mime.test.ts (accept/validation/badges/normalization). Verified: uploading a real video through the route returns synopsis + timestamped transcript/keyframes as text (curl, real AliDocMind key). Part of #886. * feat(extraction): use Alibaba Cloud icon for AliDocMind provider Replace the placeholder bailian.svg with a dedicated Alibaba Cloud icon mark (the square symbol from the official wordmark, text removed) so the provider reads as Aliyun rather than Bailian. Square aspect matches the other provider icons. Source: Alibaba Cloud / Alibaba Group brand assets. * style: prettier format AliDocMind + media extraction files * fix(extraction): harden AliDocMind — SSRF guard, cred verify, env gate, table text Addresses review findings on the AliDocMind provider: - SSRF (high): the media branch of /api/extract-document now runs the same validateUrlForSSRF check on a client-supplied baseUrl as the document branch, so an audio/video upload can't point the server's Aliyun SDK at an internal host. - Credential verify (high): verifyAliDocMindCredentials now whitelists success signals instead of blacklisting auth errors. Probed against the real API: valid creds + bogus job returns a no-throw "BizIdNotExistOrResultExpired" body; invalid creds throw InvalidAccessKeyId.NotFound. Only a no-throw response or a "biz-not-found" business error counts as valid — an unreachable endpoint, a localized error, or throttling now correctly reports failure instead of a false "connection successful". - Env-fallback gate (med): resolveCredentials no longer reads ALIDOCMIND_ACCESS_KEY_ID/SECRET unconditionally. Env fallback is opt-in via allowEnvFallback, which the route sets only for a server-managed provider, so an unauthenticated client request can't silently run on the server account. - getDocParserResult error body (med): fetchResult throws on a non-200 result envelope instead of returning {} (empty text presented as success). - Media pagination (med): stop after the first page when segments[] are present — layoutNum/layoutStepSize address layout blocks, not media segments, so re-requesting would loop over the same segments to the safety cap. - Table content (med): tables/charts carry content in llmResult, not markdownContent; the layouts→text mapping now prefers llmResult for those types so table content isn't dropped (tables: true was advertised). - Dead code: remove getCurrentMediaParseConfig (referenced non-existent store fields; the single-panel UI reuses pdfProvidersConfig). - Dedup: media MIME list lives only in lib/document/mime.ts (ALIDOCMIND_MEDIA_MIMES); the media registry imports it. - Tests: smoke test asserts durationMs is in ms (>1000 for a ~52s clip) to guard a ms/s unit mismatch; uses allowEnvFallback for the env-cred path. Verified with the real key: PDF + video smoke tests pass; verify classifies valid vs invalid creds correctly. Part of #886. * feat(extraction): extract AliDocMind images to base64 (parity with unpdf/MinerU) AliDocMind embeds figure/picture image URLs inside each layout's markdownContent (markdown ``), not a dedicated field, and the URLs are short-lived OSS signed links. Previously we emitted `images: []` and left the expiring URLs inside the extracted text. Now, for figure/picture layouts we: - parse the OSS image URL out of markdownContent, - fetch it at extraction time (before the signature expires) and re-encode to PNG base64 via sharp — the same base64 `images[]` contract unpdf/MinerU produce, so downstream storeImages → IndexedDB → slide works unchanged, - populate metadata.pdfImages + imageMapping (the generation flow prefers pdfImages), - strip the remote-URL markdown from the emitted text so expiring links don't leak into the prompt. Downloads run concurrently; a failed/`sharp` image is dropped, never failing the whole parse. Also fixes the prior over-broad table handling: only `table` blocks read llmResult; `figure` is treated as an image (chart-figure llmResult still kept in text). Verified with the real key: a sample PDF yields 24 base64 images in both images[] and metadata.pdfImages, and no oss-cn-hangzhou URLs remain in text. Part of #886. * docs(test): note AliDocMind video smoke test is non-deterministic server-side * fix(extraction): correct AliDocMind pageCount (pageNum is 0-based) Verified against a real response: AliDocMind reports pageNum 0..13 and pageCountEstimate 13 for a 14-page document — both are 0-based. The metadata pageCount previously used pageCountEstimate directly, undercounting by one. Use the already-1-based maxPage, falling back to pageCountEstimate+1 only when no blocks were seen. Document the 0-based convention at the normalization site. * fix(extraction): address AliDocMind review — verify/SSRF/config/media (P1+P2) Resolves all P1/P2 findings from the cross-review on #887. P1 (blocking): 1. verifyAliDocMindCredentials now inspects the no-throw response body.code. An OSS-only key returns NoPermission without throwing; previously that was green-lit as "connection successful" then failed at extraction. Only a success/200 or the job-not-found probe code is accepted. Deterministic mocked test added (tests/document/alidocmind-verify.test.ts). 2. verify route trust boundary: managed → server-owned AK/SK + default endpoint only (ignore client values); unmanaged → client creds only, never env fallback, and the client endpoint is SSRF-validated before signing. 3. image fetch hardened: restricted to Aliyun OSS hosts, redirects disallowed, per-image byte cap, image-count cap, bounded concurrency (was unbounded Promise.all over provider-returned URLs). 4. AliDocMind is now selectable in the generation toolbar — availability recognizes the AK/SK pair, not just apiKey. P2 (correctness): 5. Explicit server-config for the AK/SK pair (applyAliDocMindFallback + resolveManagedAliDocMindCredentials); verify and extract now resolve managed/env identically instead of verify-uses-env / extract-rejects. 6. Poll loop checks body.code before status, so a body-level error (e.g. NoPermission) fails fast instead of retrying for the full 15 min. 7. Image page numbers preserved through fetch/filter — no longer hard-coded to page 1, so multi-page image→page association is correct. 8. Empty media extraction (no synopsis/transcript/keyframes) returns 422 PARSE_FAILED instead of HTTP 200 with empty text. 9. Format matrix trimmed to the official contract: images JPG/JPEG/PNG/BMP/GIF (dropped WebP/JP2), media MP4/MKV/AVI/MOV/WMV/MP3/WAV/AAC (dropped M4A). 10. A document-only provider (unpdf/mineru) uploaded with a media file now returns a clear 4xx instead of an opaque 500. P3: credential-verify failures return INVALID_CREDENTIALS 4xx (not INTERNAL_ERROR 500); formatTimestamp emits HH:MM:SS past one hour. Verified with the real key: verify classifies valid→ok and invalid→fail; PDF (24 images, correct page numbers) and MP4 extraction pass end-to-end. * fix(extraction): make AliDocMind selectable + correct analysis label for media Two UI/UX fixes found while manually testing the AliDocMind flow end-to-end: - Persisted-state backfill: add ensureBuiltInPDFProviders so a PDF/document provider added after a user's settings were persisted (AliDocMind) is backfilled into pdfProvidersConfig on rehydrate. Without it the provider never appeared in the store, so it couldn't be selected and never picked up its server-configured flag. Wired into both persist migrate() and merge(), mirroring the existing image/video/web-search backfills. - Analysis step label: getGenerationStepText showed "解析 document 文件" for audio/video (the type map fell through to the literal "document"). Documents keep their precise token (PDF/DOCX/PPTX/XLSX/images); audio/video now use a dedicated, locale-correct string (generation.analyzingMediaMaterial, added to all 8 locales) instead of forcing a format token into the "{{type}} 文件" template. Manually verified end-to-end (real key): PDF, MP4, and MP3 (audio extracted from the sample video) each extract and drive full course generation; a DocMind-restricted AK/SK (OSS-only) now correctly fails verification with NoPermission (400) instead of a false "connection successful". * fix(extraction): address 2nd-round AliDocMind review (managed creds, stream cap, empty verify) Resolves the three follow-up findings on #887: 1. [P1] YAML-managed AliDocMind creds now reach extraction. Both extract paths (document + media) previously cleared the managed AK/SK and relied on an env-only fallback, so a YAML-only deployment verified but failed to extract. They now resolve server-owned creds via the shared resolveManagedAliDocMindCredentials() (env OR YAML), matching the verifier. Regression test added for YAML-only creds with no ALIDOCMIND_* env. 2. [P1] Image download is now size-capped while streaming. Instead of buffering the whole response and checking length afterward, the body is read chunk by chunk with a cumulative byte count that aborts the moment it exceeds the cap — a missing/false Content-Length can no longer exhaust memory. Host allowlist tightened to oss-*.aliyuncs.com. Tests: non-OSS host refusal, no-Content-Length overflow abort, declared-oversize rejection. 3. [P2] Empty verification body no longer counts as success. Removed the codeStr === '' branch from the positive-signal whitelist (a working key always returns the job-not-found business code for the bogus probe id). Regression test added for empty and absent bodies. Rebased onto main (package.json: kept @openmaic/storage + AliCloud deps). Verified with the real key: PDF + MP4 extraction still pass end-to-end. * fix(extraction): merge AliDocMind AK/SK into a baseUrl-configured YAML entry Follow-up to the YAML-managed credential fix. When a YAML `pdf.alidocmind` entry also specifies `baseUrl`, the generic loadEnvSection() (pdf requires a baseUrl) creates the pdf.alidocmind entry copying only apiKey/baseUrl/models/ proxy — never AK/SK. applyAliDocMindFallback() then returned early because the entry already existed, so the provider was "managed" but had no usable credentials, and resolveManagedAliDocMindCredentials() returned undefined (verify + extract both silently lost the creds). Merge the AK/SK into the existing entry instead of returning early. Added a regression test with baseUrl + accessKeyId + accessKeySecret together (the previous test omitted baseUrl, so the generic loader skipped the entry and the bug was masked). * fix(extraction): don't mark AliDocMind managed without AK/SK; align poll code check Two edge cases from a follow-up adversarial pass: - [MED] A YAML `pdf.alidocmind` entry with `baseUrl` but no AK/SK (and no env AK/SK) made the generic loader create the entry → isServerConfigured=true (managed) → but resolveManagedAliDocMindCredentials() returned undefined, so the provider was locked out AND client-entered AK/SK were silently dropped. applyAliDocMindFallback now deletes a credential-less entry so the provider stays UNMANAGED (clients supply their own creds). Regression test added. - [LOW] verify accepted a `code: "success"` body but the extraction poll loop threw on any non-"200" code — a success-shaped status would pass verification then fail extraction. The poll loop now treats "200"/"success" as benign, matching verifyAliDocMindCredentials. --------- Co-authored-by: wyuc <wang-yc24@mails.tsinghua.edu.cn> |
||
|
|
dd83ac8669 |
feat(pbl): read flip — fold learner state from RuntimeStore with snapshot backfill (#869 Part C, step 3) (#902)
* feat(pbl): read flip — fold learner state from RuntimeStore with snapshot backfill (#869 Part C, step 3) Step 3 of #869 Part C: on stage load, PBL learner state is reconstructed by folding RuntimeStore records over the design-time template. Records are made self-sufficient at drain time (id-only outbox events gain content attachments resolved from the in-hand project), pre-existing and ring-evicted history is recovered through snapshot records, and the document stays authoritative whenever the fold disagrees — with a self-heal snapshot so the next load converges. The #893 dual-write and document persistence are untouched; removing them is the next step. - lib/pbl/v2/runtime/learner-state.ts: PBLLearnerState split with extract/apply/strip — codifies the document/runtime boundary - lib/pbl/v2/runtime/record-payloads.ts: enriched event, engagement and snapshot record payload shapes (app-owned, versioned) - lib/pbl/v2/runtime/fold.ts: seq-ordered, id-deduped, epoch-aware fold with gap diagnostics; snapshots supersede earlier gaps - lib/pbl/v2/runtime/hydration.ts: catch-up drain, fold, shadow compare, document-authoritative fallback, deduped self-heal snapshots; never throws into the stage load path - thread_compacted runtime event: instructor memory compression was a learner-state mutation with no event; deterministic id from the post-compaction thread shape - equivalence property test: randomized reducer runs into a real BrowserRuntimeStore must fold back to extractLearnerState(project) Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(pbl): harden read-flip fold and hydration after cross-review - remove thread_compacted: instructor compaction mutates a per-request server copy after the last streamed patch, so neither the compacted thread nor the event ever reached the persisted client document — earlierSummary stays extracted for forward-compat only - fold: apply the shared engagement ring cap; clear gaps on project_reset (matching the snapshot branch); default-branch gap + compile-time exhaustiveness for future event kinds - hydration: preserve document-only transient fields (runtimeEvents, pendingOpenTaskPriorQuizResults) on fold-match; lodash isEqual instead of JSON.stringify equality; hydrate scenes in parallel; classroom server-fallback load path now hydrates too - stage load: re-check the in-memory guard after async hydration so a concurrent setScenes is not clobbered by the older disk snapshot - drain: serialize overlapping drains per (stage, scene, learner) so duplicate records are not appended - drop write-only anchor.recordSeq and the 'test' reason variant; shared attachment-required kind set; single clone helper; strict enriched-payload assertion; learner-state field inventory test Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(pbl): close stale-load races and drain head-of-line blocking - loadFromStorage bails when a different stage appears during the now-async hydration window, not just the same stage - classroom server fallback re-checks stage identity after hydration before applying scenes - each queued drain is individually time-bounded so a hung append cannot block the per-key drain chain forever - unknown forward-version event kinds enrich into a valid payload (attachment null + reason) instead of a bare event Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(pbl): address review — no empty-scenes snapshot, truly serialized drains - classroom fallback: hydrate first, then apply stage + scenes in the same synchronous update, so the debounced save can never persist { stage, scenes: [] } while hydration is in flight - drain: the per-key chain now awaits the actual drain work (never the timeout race), and the work checks a cooperative deadline before each append and before watermark persistence — callers keep the bounded wait, overlapping same-key appends are impossible Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(store): only abort a load when navigation happens during hydration The post-hydration guard compared the resident stage against the load target, so a load started while another stage was legitimately resident (direct loadFromStorage with no prior setStage) bailed out before the legacy generationComplete self-heal ran. Snapshot the stage id at entry and abort only when it changed to a different stage during the await. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(pbl): release a wedged same-key drain slot after a hard cap The serialization chain linked on the raw work promise, so a single appendRecord that never settles held the per-key slot forever. The chain link now races the work against a hard cap (2x the drain timeout): the cooperative deadline already prevents an expired drain from appending further records or touching the watermark, so releasing the slot lets the successor proceed while the worst case from the straggler is one id-dedupable in-flight record. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(store): serialize scene-committing loads with a monotonic request token Guard predicates over resident stage ids cannot order overlapping loads from the same origin (load C then load D from resident B — the later navigation lost). Every async scene-committing load now claims a shared monotonic token at entry and commits only while still the newest request; setStage and clearStore advance the token so sync navigation invalidates pending loads. The classroom fallback shares the same counter, so both entry points order against each other. The same-stage in-memory scenes recheck stays: generation's setScenes is not a tokened load. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(classroom): bind the server fallback to its navigation's load token Claiming a fresh token at apply time laundered an old request into the newest one: a stale /api/classroom response could overwrite a newer navigation. The classroom page now claims one token at load start and threads it through both the IndexedDB attempt and the server fallback; the fallback helper only validates that origin token at commit. loadFromStorage accepts the pre-claimed token, claiming internally for existing callers. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(classroom): cancel stale classroom loads across every async phase The load-token guard only protected the scene commit; the rest of the page load flow ran to completion for a superseded navigation, so a stale run could merge the old stage's media tasks back into the shared map, clear and repopulate the generated-agent registry over the newer classroom, and drop the newer navigation's loading state. Extract the orchestration into runClassroomLoad with an isCurrent() gate — the effect-cleanup cancel flag combined with the load token — checked after every await and before every shared-store write (restore, agent save/load, settings, media kickoff, setError/setLoading). A superseded run now performs zero further side effects. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(classroom): do not invalidate the current load token when committing fallback scenes The fallback commit went through setStage(), which claims a fresh scene-load token — so a legitimate fallback load invalidated its own token, the post-commit isCurrent() check then returned false, and the guarded finally skipped setLoading(false), leaving the page stuck on the loading screen (E2E never saw the sidebar/rail). Commit fallback stage+scenes via a token-free setState (applyClassroomStageAndScenes) instead. Stale loads are still blocked by the token checks before the apply; only the load's own commit no longer cancels itself. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(classroom): resume generation for the current classroom regardless of load token The auto-resume generation effect was gating generateRemaining / markGenerationCompleteIfDone / generateMediaForOutlines on the scene-load token (via activeLoadTokenRef). By the time the effect ran the token had advanced, so generation for the freshly-mounted classroom was silently skipped — a deck arriving from generation- preview with pending scenes never finished, leaving the Edit-course toggle disabled (the two slide E2E specs timed out clicking it). Resuming generation for the mounted classroom is not a competing scene-load; navigation away is already handled by the effect's own cancelled flag (which also stops()). Gate resume on cancellation only. The load token still guards scene-committing loads (runClassroomLoad, fallback apply, loadFromStorage), so the review-round race fixes stay. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(classroom): stop cancelling generation resume on benign effect re-runs The generation-resume effect gained a cancelled-flag guard around generateRemaining / markGenerationCompleteIfDone / generateMediaForOutlines. generateRemaining has a new identity every render (useSceneGenerator's useCallback depends on the options object, and the classroom passes a fresh { onComplete } literal each render), so the effect re-runs on every render and its cleanup trips cancelled. The kickoff runs inside an async loadImageMapping().then(...), which resolves after a re-run has set cancelled, so generation was silently skipped — the deck never left 'Generating...', the Edit-course toggle stayed disabled, and two slide E2E specs timed out. Restore the generation-resume effect to its prior unguarded form (matches origin/main): navigation-away already stops the generator via the load effect's stop() cleanup. The scene-load-token guards on the actual load/commit paths (runClassroomLoad, fallback apply, loadFromStorage) are untouched, so the review-round races stay fixed. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(classroom): make generated-agent replacement atomic * fix(classroom): clean up stale restored media * fix(pbl): harden learner-state boundary * fix(pbl): serialize hydration side effects * fix(pbl): bound hydration drain barrier * fix(pbl): bound queued hydration drains --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com> Co-authored-by: 杨慎 <117187635+cosarah@users.noreply.github.com> |
||
|
|
c56929510c |
feat(web-search): add SearXNG provider support (#842)
* feat(web-search): add SearXNG provider support Enable self-hosted SearXNG via SEARXNG_BASE_URL, with server-side provider fallback, settings sync, and tests. Co-authored-by: Cursor <cursoragent@cursor.com> * style: fix Prettier formatting in web-search files Restore pnpm check by applying Prettier to imports, headers, and test assertions. Co-authored-by: Cursor <cursoragent@cursor.com> * fix(web-search): keep SearXNG base URLs server-managed only Close the SSRF primitive by ignoring client-supplied SearXNG URLs at the route boundary and rejecting them in resolveSafeClientWebSearchBaseUrl. Add negative route/config tests and an opt-in live smoke test for the JSON API. Co-authored-by: Cursor <cursoragent@cursor.com> * style: fix Prettier formatting for SearXNG web-search changes Co-authored-by: Cursor <cursoragent@cursor.com> --------- Co-authored-by: 徐松 <song.xu@aieshanghai.com> Co-authored-by: Cursor <cursoragent@cursor.com> |
||
|
jackefnandRowan_lxb
|
baff0998bb |
Feat/document bundles milestone 3 (#844)
* feat(document): support document bundles * fix(document): avoid server extractor imports in client bundle * fix(pdf): default mineru backend to pipeline * fix(document): address bundle review findings * fix(document): harden extraction limits and storage errors * fix(document): make analysis step format agnostic --------- Co-authored-by: Rowan_lxb <Lxb_savior@163.com> |
||
|
wyucand杨慎
|
ce89b4c314 |
fix(home): clarify Interactive Mode selected state (#901)
* fix(home): clarify interactive mode selected state * fix(home): soften light interactive mode state * fix(home): honor reduced motion in dark mode * test(home): assert current interactive mode styles --------- Co-authored-by: 杨慎 <117187635+cosarah@users.noreply.github.com> |
||
|
|
b4c1c580f8 |
fix(generation): localize scene generation errors (#894)
Co-authored-by: wsun1 <wsun1@users.noreply.github.com> Co-authored-by: wyuc <wang-yc24@mails.tsinghua.edu.cn> |
||
|
PhillLittlewoodandwyuc
|
a2886af92a |
Added ComfyUi to has Image Provider (#850)
* feat: integrate comfyui workflows into image provider selection * style: format code with prettier * fix: align i18n keys for russian locale * feat: implement requested changes and format code * feat: implement requested changes including renaming default workflow --------- Co-authored-by: wyuc <wang-yc24@mails.tsinghua.edu.cn> |
||
|
Yanpeng Wang
|
b5e1963e48 |
feat(document): rename to Document Parsing, show supported formats, extend MinerU support
Squashed from PR #837 after review approval and green CI. |
||
|
Yanpeng Wang
|
d54e62f5c4 |
feat(token-plan): one-click token-plan setup + deployment usage dashboard (#784)
* feat(usage): add usage normalization, pricing, model-fetch, balance, storage Foundational layer for token-plan usage tracking (cc-switch-modeled): - lib/usage/normalize.ts: AI SDK v6 usage → four-class token shape - lib/usage/pricing.ts + defaults: per-class USD pricing table - lib/server/model-fetch.ts: /models candidate-URL multi-fallback (ported) - lib/usage/balance-providers.ts: built-in balance queries + detection - lib/server/usage-storage.ts: fire-and-forget jsonl logging to data/usage All pure/storage logic covered by vitest (32 tests). * feat(usage): capture token usage at the callLLM/streamLLM chokepoint - callLLM records result.usage before returning - streamLLM wraps onFinish to record totalUsage on stream completion, preserving any caller-supplied onFinish - provider/model derived from the model instance (no route changes) - fs-backed storage imported dynamically; fire-and-forget, never throws - include_usage is already sent by @ai-sdk/openai, so streaming is covered * feat(usage): add probe-models, balance, and usage API routes - POST /api/provider/probe-models: discover chat models via /models with candidate fallback; SSRF-guarded; filters non-chat ids; 401/404 typed - POST /api/provider/balance: built-in balance detection + billing fallback - GET /api/usage: aggregate jsonl by model/day/source with costIncomplete flag Verified e2e against the live MAIC gateway: 16 chat models (6 filtered), balance detected, and a real callLLM writes a costed usage row. * feat(settings): add token-plan preset picker to provider dialog - lib/config/token-plan-presets.ts: data-driven vendor presets (Huawei/MiniMax/ Xiaomi token plans, OpenRouter/SiliconFlow gateways, DeepSeek/GLM/Qwen/Hunyuan/ Doubao direct) with baseURL, protocol, optional modelsUrl, category - add-provider-dialog: '选择厂商/自定义' tabs; picking a preset auto-fills baseURL+protocol+modelsUrl, custom tab unchanged - ProviderSettings.modelsUrl carries the optional /models override - i18n keys added across all 8 locales Verified in-browser: preset picker renders grouped by category. * feat(settings): add Fetch Models button and balance bar to provider panel - 拉取模型: probes /models, merges discovered ids into the model list (dedupe, keeps manual additions), with success/no-endpoint/auth messages - 查询余额: queries /api/provider/balance, renders a balance bar or a 'check console' hint when unsupported - index.tsx: handleModelsFetched merges probe results into provider config - i18n keys across all 8 locales; removed a now-unused eslint-disable Verified in-browser against MAIC gateway: 16 models fetched, balance shown. * feat(settings): add usage dashboard to System Settings - usage-dashboard.tsx: echarts dual-axis daily trend (tokens + cost), totals cards, by-model table, refresh; reads GET /api/usage - mounted at the top of GeneralSettings (系统设置) - honest disclaimer + costIncomplete marker when a model lacks pricing - i18n keys across all 8 locales Verified in-browser: shows 1 request / 31 tokens / $0.0003 from a prior call. * feat(token-plan): multi-modal one-click setup in System Settings - token-plan-presets.ts: presets now declare per-modality targets (llm/image/video/tts/webSearch); MiniMax is the full-set template, others LLM-only — extend by adding entries (at-our-best adaptation) - apply-token-plan.ts: fills one key into every declared modality via injected store setters, isolating per-modality failures (TDD, 4 tests) - token-plan-settings.tsx: new sidebar page — pick plan, enter key, one-click apply lights up adapted modalities (+LLM model probe), shows 'not adapted yet' for the rest, balance bar reused - reverted add-provider dialog to plain custom form (preset picker moved here) - i18n across all 8 locales Verified in-browser: Token Plan page renders, MiniMax shows all 5 modalities, apply/balance UI wired. 859 tests pass, build clean. * feat(token-plan): add custom token plan entry to Token Plan page - Custom card at the bottom of the provider list: pick it to manually enter name + protocol + baseURL (LLM-only), then key + one-click apply via the same applyTokenPlan flow (mirrors cc-switch's custom provider) - effectivePreset unifies preset vs custom for apply/balance/probe - i18n keys (customGroup/customName/customHint) across all 8 locales Verified in-browser: custom card expands the manual form; preset flow intact. * feat(token-plan): reflect persisted config on the Token Plan page Other settings panels read the store directly, so they survive section switches; the Token Plan page only wrote the store, so it looked blank on return. Now it also reads providersConfig: - selecting a preset prefills its saved API key - configured presets show a '已配置/Configured' badge State persistence was never broken (apply writes zustand→localStorage); this fixes the missing read-back so the page reflects it. * fix(settings): isLLMProviderConfigured crashed on providers without models Root cause of 'token plan config disappears': applyTokenPlan writes a new LLM provider with no models yet (probe fills them later). isLLMProviderConfigured did config.models.length unguarded → threw inside setProviderConfig's resolver → the whole set() aborted → key/baseUrl/type never persisted. - guard models in isLLMProviderConfigured (shared validator; affects any provider written without a models array) - seed models:[] in applyTokenPlan's LLM write for a valid initial shape - regression tests: validator no longer throws; apply+probe write keeps apiKey Verified in-browser: DeepSeek persists key and shows 已配置 after section switch. * feat(token-plan): add remove/teardown for a configured token plan Apply had no inverse. Add removeTokenPlan: clears the API key and disables (enabled:false) every modality the plan declared; a custom LLM provider is deleted entirely (removeProvider), built-ins keep the cleared shape. - removeTokenPlan in apply-token-plan.ts (mirrors applyTokenPlan, isolated per-modality, injected setters; 3 tests) - trash button on configured preset cards in the Token Plan page; resets page state if the removed plan was selected - i18n 'remove' across all 8 locales Verified in-browser: removing DeepSeek empties its key and drops the 已配置 badge. * feat(usage): track multimodal usage, drop cost/pricing Reframe usage stats as pure usage (no cost), per user decision: - usage-storage: drop all cost fields; add kind (llm/image/video/tts/asr) + quantity + unit; LLM keeps token counts, others store quantity - delete lib/usage/pricing.ts + pricing-defaults.json + its test - instrument image/video/tts at the server-only API routes (not in the provider dispatch — those files are in the client graph and importing fs-backed usage-storage broke the client bundle) - /api/usage: aggregate by model/day/modality, no cost - dashboard: per-modality usage (tokens/images/seconds/chars), token-only daily trend, no '$'; i18n cost keys replaced with modality/unit keys - backward compatible: legacy rows (no kind, stray cost fields) read as llm 912 tests pass, build clean. * feat(usage): per-modality dashboard layout + softer dark-mode chart - group usage into per-modality sections (LLM/image/video/tts/asr), each table's usage column uses one consistent unit (token/image/sec/char) — no more mixed units in a single column - summary chips per modality with their own unit - trend chart now plots daily REQUESTS (unit-agnostic, works for any modality) instead of LLM-only tokens - theme-aware chart: faint thin line + soft gradient area, muted axis/grid colors via useTheme — fixes the harsh solid stroke in dark mode Verified in-browser (dark): TTS shows '字符', LLM shows 'Token', separate sections; chart no longer has a hard line. * refactor(usage): dedupe model/fetch/usage helpers, parallel balance probe Review cleanups on the token-plan/usage branch: - extract modelInfoFromId() (shared vision heuristic + ModelInfo shape) - extract fetchWithTimeout() shared by model-fetch and balance-providers - extract recordGenerationUsage() to dedupe the image/tts/video routes - queryBalance: fetch billing subscription + usage in parallel - parseOneApiBilling: report quota without remaining when usage endpoint is unavailable, instead of implying zero spend (full balance) - split the two jammed imports in settings/index.tsx * feat(token-plan): drop custom token-plan support Custom token plans (manual baseURL/protocol entry) added complexity for little gain — a one-off provider is better configured directly on the Providers page. Token Plan is now preset-only: - remove custom mode, manual fields, and the custom card from the UI - collapse effectivePreset back to the selected preset - drop the now-dead removeProvider action + custom-id branch in removeTokenPlan - remove orphaned customGroup/customName/customHint i18n keys (8 locales) * feat(token-plan): add Volcengine/Tencent/Bailian plans, drop balance feature Add three vendor token-plan presets (all map to existing built-in LLM providers, so it's data-only — no new adapters): - 火山方舟 Volcengine Ark → doubao, OpenAI /api/v3 - 腾讯 TokenHub Token Plan → tencent-hunyuan, OpenAI /plan/v3 (the plan-specific base; /v1 is the pay-as-you-go gateway) - 阿里百炼 Token Plan → qwen, cross-model plan (Qwen + DeepSeek/Kimi/ GLM/MiniMax) on one key; model list is probed/entered Remove the balance/quota feature entirely — we now track usage, not cost, and every vendor's balance query needs its own cloud AK/SK + signature (Volcengine SigV4 / Tencent TC3 / Aliyun BSS), which the Bearer-key billing-endpoint probe never supported anyway: - delete lib/usage/balance-providers.ts, /api/provider/balance, its test - strip the Check Balance button + balance bar from the token-plan page and the provider config panel - remove the 4 balance i18n keys across all 8 locales - restore an eslint-disable the branch had dropped in provider-config-panel * feat(token-plan): use cloud-brand logos for vendor token plans The three vendor plans are cloud offerings, not single-model products, so icon them with the cloud brand rather than a model logo: - 火山方舟 → volcengine.svg (was doubao.svg) - 腾讯 TokenHub → tencentcloud.svg (was hunyuan.svg) - 阿里百炼 → alibabacloud.svg (was bailian.svg) Logos are the colored brand variants from lobehub/lobe-icons, matching the existing colored-logo style (plain <img>, no dark:invert needed). * feat(token-plan): keep only MiniMax and Volcengine presets Trim the token-plan list to the two we want to ship: MiniMax (full-set template) and 火山方舟 Volcengine Ark. Drop the Tencent/Bailian plans and the OpenRouter/SiliconFlow/DeepSeek/GLM/Qwen entries. - remove the now-unused tencentcloud.svg / alibabacloud.svg logos (volcengine.svg stays; the other logos are still used by the provider registry) - retarget the LLM-only apply test from the deleted deepseek preset to volcengine-ark * feat(token-plan): restore aggregator/third-party presets Previous commit over-trimmed: the intent was to drop only the Tencent and Bailian token plans, not the OpenRouter/SiliconFlow/DeepSeek/GLM/Qwen entries. Bring those back; keep only Tencent/Bailian removed. - token_plan: MiniMax, 火山方舟 Volcengine Ark - aggregator: OpenRouter, SiliconFlow - third_party: DeepSeek, GLM, Qwen Revert the apply test back to the deepseek fixture (restored). tencentcloud.svg / alibabacloud.svg stay deleted (their plans are gone). * fix(token-plan): point Volcengine plan at the Coding Plan endpoint The plan's ark--prefixed API keys authenticate only against /api/coding/v3, not the general /api/v3 endpoint — the latter rejects them with "The API key format is incorrect", so model probing returned nothing. Switch the base URL to https://ark.cn-beijing.volces.com/api/coding/v3. * fix(token-plan): Volcengine is an Agent Plan (Anthropic /api/plan) Per the Ark Agent Plan docs, the ark--prefixed keys authenticate ONLY against the dedicated Anthropic-compatible base https://ark.cn-beijing. volces.com/api/plan ("其他 Base URL 无法在 Agent Plan 中使用"). The general /api/v3 and the Coding Plan /api/coding endpoints both reject the key as "API key format is incorrect", which is why model probing kept returning 0. - baseUrl → https://ark.cn-beijing.volces.com/api/plan/v1 (the /v1 lets the Anthropic SDK land on /api/plan/v1/messages) - apiFormat → anthropic - rename to 火山方舟 Agent Plan Probe still targets /api/plan/v1/models (the path exists); if the Anthropic gateway doesn't return an OpenAI-shaped list, users fall back to typing a model id like ark-code-latest. * fix(token-plan): Volcengine Agent Plan = OpenAI /api/plan/v3 + ark-code-latest Settled after probing the real key and reading cc-switch's approach: - The ark- plan key works on the OpenAI-compatible /api/plan/v3 endpoint (chat/completions returns 200); switch apiFormat back to openai. - The plan exposes NO /models list (every /api/plan/*/models is 404), which is why probing kept returning 0. cc-switch handles this by hardcoding a single ark-code-latest (an auto-routing alias valid on any tier) and does NOT use AK/SK for model listing — so we do the same. - Seed defaultModels: ['ark-code-latest'] only; users add specific ids by hand. Supporting machinery (kept, general-purpose): - applyTokenPlan seeds models from defaultModels instead of wiping to [] - handleApply uses defaultModels and skips the doomed probe when present - drop stray .playwright-mcp/ debug artifacts and gitignore them * style: fix prettier formatting in usage files CI runs prettier on the whole repo (prettier . --check); these four files predate this branch's formatting pass and tripped the check. * feat(token-plan): verify Volcengine Agent Plan's published model set The Agent Plan publishes a fixed model set but exposes no /models endpoint, so carry the documented models as CANDIDATES and verify each on apply: - add verifyModels flag to TokenPlanModalityTarget - new /api/provider/probe-chat-models route: sends a minimal chat request per candidate (OpenAI /chat/completions or Anthropic /messages) in parallel, returns the subset that succeeds; SSRF-guarded, auth-failure short-circuits - handleApply gains a verify branch (before the fixed-defaultModels fast path), falling back to the seeded list if verification fails - Volcengine preset now carries the 12 published Agent Plan text models (doubao-seed-2.0-*/deepseek-v4-*/minimax-m*/glm-5.2/kimi-k2.*) as candidates This auto-prunes retired (docs flag deepseek-v3.2/glm-5.1 as 即将下线) and tier-gated models without code changes. Verified all 12 resolve against a real plan key. * feat(token-plan): wire Volcengine Agent Plan image + video modalities Make the Ark seedream/seedance adapters path-configurable and light up the image/video modalities on the Volcengine plan: - seedream/seedance adapters: resolveArkRoot() uses baseUrl verbatim when it already carries an /api/... path (token plan's /api/plan/v3), else appends the standard /api/v3 — no regression for the pay-as-you-go default host. - applyTokenPlan: image/video branches inject a modality's defaultModels as customModels and set them as the active provider+model, so generation works out of the box. New optional setImageProvider/ModelId + setVideoProvider/ ModelId actions (UI passes the store setters; tests omit them). - Volcengine preset declares image (doubao-seedream-5.0-lite, verified 200 on /api/plan/v3/images/generations) and video (doubao-seedance-2.0/1.5-pro — Medium+ tiers only; lower tiers reject at call time, no code change needed to upgrade). Applying the plan overwrites the shared seedream/seedance slot with the plan config (same overwrite model as LLM); switching back to pay-as-you-go is a manual edit or plan removal. Verified image end-to-end with a real plan key. * feat(token-plan): verify image/video models on apply, disable unsupported tiers The Volcengine plan lit up video optimistically, but lower tiers (Small) don't include video — so using it 404'd with UnsupportedModel. Probe media models on apply and only keep what the tier actually supports: - generalize /api/provider/probe-chat-models with a `kind` (chat|image|video): image hits /images/generations, video hits /contents/generations/tasks with empty content. The model-support check (404 UnsupportedModel) runs before any billable work, so probing never starts a real image/video job; for media, "supported" = any non-404 response. - handleApply: after lighting up image/video, probe each verifyModels modality; prune to the verified model set + re-select a working model, or disable the modality entirely if none pass (no false "available"). - Volcengine preset: image/video targets gain verifyModels: true. - add settings.tokenPlan.tierUnsupported across 8 locales. Verified with a real Small-tier key: image (seedream-5.0-lite) passes and is kept; video (seedance-2.0/1.5-pro) 404s and is disabled. * feat(web-search): add Doubao (豆包搜索) provider Doubao Search (Custom 版) over its REST endpoint POST open.feedcoopapi.com/search_api/web_search with Bearer auth — the same endpoint the askecho-search-infinity MCP server wraps, so the Volcengine Agent Plan key authenticates directly. Mirrors the MiniMax adapter: maps Result.WebResults to WebSearchSource (prefers Summary, the query-relevant excerpt, over Snippet for LLM use) and surfaces errors from ResponseMetadata.Error. - register 'doubao' in WebSearchProviderId + WEB_SEARCH_PROVIDERS - searchWithDoubao adapter, searchWeb dispatch, store default config - SSRF allowlist entry for the search host * feat(audio): support Agent Plan single-key auth for Doubao TTS generateDoubaoTTS now picks auth + endpoint from the key shape, since Volcengine exposes Seed-TTS as two products with separate credentials (verified: a plan key 401s on the normal endpoint, and the plan endpoint rejects appId-style auth): - single key (no colon) -> X-Api-Key, for the Agent Plan /plan endpoint - appId:accessKey -> X-Api-App-Id + X-Api-Access-Key (unchanged) A malformed pair (empty half) fails clearly instead of sending an empty header. Reuses the existing NDJSON/base64-mp3 parsing and voice list. * fix(media): map MiniMax video 720p to its real 768P tier normalizeVideoOptions defaults minimax-video to '720p' (the first supported resolution), but Hailuo 2.3 only accepts 768P/1080P and rejects 720P with '2013 ... does not support resolution 720P'. MiniMax's mid tier is 768P, not 720P (the adapter already falls back to 768P, as does the connectivity test), so map the shared enum's '720p' to 768P. Regression tests lock the mapping. * feat(token-plan): add web search + TTS to Volcengine Agent Plan, widen image tiers Extend the volcengine-ark preset now that the adapters exist: - webSearch -> doubao (own host open.feedcoopapi.com, not the ark endpoint) - tts -> doubao-tts on the /api/plan/tts endpoint (single-key auth) - image defaultModels widened to a best-first Seedream 5.0/4.5/4.0 list so a higher tier keeps the strongest model while verifyModels prunes the rest; video keeps the 2.0 + 1.5-pro candidates Comments record the verified host/auth quirks of each modality. * feat(token-plan): show result panel only after probing, with two clear states Addresses review feedback that a green check implied generation works when it only meant 'configured'. The panel now renders after probing finishes (gated on results && !applying) so it reflects the final set, and uses two states: green when the modality is configured/usable, muted when a live probe proved it unavailable (e.g. video on a tier without it). * feat(token-plan): scope presets to true multi-modal token plans Drop the single-modality LLM presets (OpenRouter, SiliconFlow, DeepSeek, GLM, Qwen) from Token Plan. A token plan's defining trait is one key spanning many modalities; those entries are ordinary LLM API providers already covered by the add-provider flow, and listing them here muddied the 'one key, every modality' promise. Only MiniMax and the Volcengine Ark Agent Plan remain. The UI already hides categories with no entries. apply-token-plan's LLM-only test now uses a local fixture instead of the removed deepseek preset. * feat(token-plan): progressive reveal of probe results on apply The result panel previously rendered all at once after probing finished, reading as dead air during the model probe. Now rows appear immediately on Apply: modalities with a live probe in flight show a spinner ('pending') and resolve to lit/failed independently as each probe returns, while non-probe modalities show lit right away. Probes run in parallel (Promise.all) instead of sequentially. A row only turns green once its own probe confirms, so this reveals structure + live progress without a premature green — complementing the earlier 'render only after probing' intent rather than reverting it. Per review feedback from @wyuc on #784. * fix(token-plan): enrich seeded models with built-in thinking capability Token Plan built ModelInfo objects from probed ids via modelInfoFromId(), filling only streaming/tools/vision — so a model that supports configurable thinking lost capabilities.thinking and InlineThinkingControl was hidden. modelInfoFromId now takes an optional providerId and overlays the catalog thinking capability for that (provider, model) pair; applyTokenPlan does the same for its synchronously-seeded list. Added the Ark Agent Plan's dotted aliases to the metadata table: - native Doubao Seed 2.0 family (doubao-seed-2.0-pro/code/lite/mini) - cross-vendor models the plan serves through its OpenAI-compatible endpoint (deepseek-v4-pro/flash, glm-5.2, kimi-k2.7-code/k2.6, minimax-m3/m2.7, ark-code-latest) All verified against a live plan key: each accepts the gateway's unified reasoning_effort field (low/medium/high) and actually reasons. They share the doubao effort adapter, which disables via 'minimal' (not 'none') — matching what the plan endpoint accepts (it rejects reasoning_effort:'none'). Addresses review point #1 from @wyuc on #784. * Improve token plan capability setup UI * fix token plan setup flow * chore: prettier format tts-providers.ts |
||
|
jackefnandYanpeng Wang
|
3b7a0ca0c0 |
feat(document): add multi-format course material upload (#741)
* feat(document): add multi-format course material upload * fix(document): support office uploads with mineru cloud * fix(document): handle mineru cloud office edge cases * fix(document): align mineru cloud artifact metadata --------- Co-authored-by: Yanpeng Wang <yanpg.wang@gmail.com> |
||
|
|
cec3c6aaea |
Refactor widget actions into the scene action pipeline (#796)
* refactor: align widget actions with scene pipeline * feat(edit): migrate legacy teacherActions off interactive content on load The widget-actions pipeline no longer persists a teacherActions authoring layer; playback reads only scene.actions. Add migrateInteractiveContent to strip the dead field from existing documents as they load through migrateScene (wired into all store/classroom load paths). Satisfies the #794 forward-migration acceptance item. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * fix(generation): default widget_setState.state to {} when the LLM omits it The removed TeacherAction->Action converter always defaulted state to {}. Restore that guard in the structured-output parser so an action missing state can't forward state: undefined into the widget iframe, where SET_WIDGET_STATE handlers dereference it. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * fix(edit): also strip nested widgetConfig.teacherActions in migration Legacy WidgetConfig variants each declared their own teacherActions, so existing documents can carry the dead field nested inside content.widgetConfig as well as at the top level. Extend migrateInteractiveContent to drop both, preserving all other widgetConfig fields. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * feat(prompts): document stable widget-action selectors for diagram/game/viz3d The interactive-actions prompt documented stable target selectors only for procedural-skill and simulation widgets. Add the canonical selectors the diagram / game / visualization3d content prompts actually emit (node IDs via revealOrder; #zoom-in-btn / #canvas-container / #speed-slider / ...; #start-btn), plus a rule to use widget_setState or a speech-only beat instead of guessing a target when no stable selector is known (e.g. free-form code widgets). Hardens against widget actions silently no-oping on non-existent selectors. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com> Co-authored-by: 杨慎 <117187635+cosarah@users.noreply.github.com> |
||
|
Yizuki_Ame
|
243e9f3a48 |
feat: auto-retry transient failures during scene generation (#788)
Adds automatic retry with exponential backoff for scene generation (content/actions/TTS): shared withGenerationRetry helper at the fetch boundary, smart error classification (retries 429/5xx/timeouts/network, fails fast on 401/403/404), AI SDK RetryError unwrapping, abort-aware backoff, per-route thinking config preserved, and upstream HTTP status surfaced for client retry classification. Fixes #755. |
||
|
杨慎andwyuc
|
0f44e9a46f |
Add PBL v2 runtime APIs and classroom UI (#799)
* feat: add PBL v2 runtime UI * fix: ship PBL v2 instructor avatar * fix: ship PBL v2 workspace logo * fix: prevent scenario PBL legacy fallback * chore: restore package lint comments * test: cover ordinary PBL legacy fallback * fix: front-load PBL single-call validation * fix: preserve PBL v2 content when assembling scenes --------- Co-authored-by: wyuc <wang-yc24@mails.tsinghua.edu.cn> |
||
|
杨慎
|
09863af421 |
Add PBL v2 core schema and generation path (#795)
* feat: add PBL v2 core planner * chore: restore unrelated lint comments * fix: document PBL v2 scenario outline flags * fix: expose PBL scenario subtype in outlines * chore: clear PBL v2 completion stats lint warning * chore: address PBL instructor code quality comments * fix: address PBL v2 review findings * chore: restore unrelated lint comments |
||
|
wyuc
|
7cb129183d |
chore(packages): publish the @openmaic/* SDK family to npm (#778) (#780)
* chore(packages): publish the @openmaic/* SDK family to npm (#778) Prepares the @openmaic/{dsl,renderer,importer} family for its first npm publish, and moves the SDK packages onto the @openmaic scope. Why the scope move: the @maic org name is unavailable on npm (an unscoped `maic` package already holds the name), so @maic/* is not claimable. @openmaic matches the project name, the scope is free, and the repo already ships an @openmaic/docs package — so the SDK family now lines up with that convention. - rename @maic/{dsl,renderer,importer} -> @openmaic/* across packages, the workspace glob, the package dir, and all import sites; lockfile regenerated - renderer: add publishConfig (public, registry.npmjs.org) — was missing, so a scoped publish would default to the wrong registry / restricted access - importer: add a files allowlist (dist, README, LICENSE) and drop the fragile .npmignore blacklist that shipped src; add an exports map so ESM consumers resolve dist/index.js instead of falling back to the .cjs main - all three: add a prepublishOnly build (+ test/typecheck) guard so a publish can never ship a stale or empty dist - add a tag-triggered publish workflow with npm provenance, pinned by name to the three @openmaic packages so the vendored forks (mathml2omml, pptxgenjs) are never published Refs #778, #720 (Phase 1). * fix(packages): address cross-review on the @openmaic publish prep Cross-review (Claude /code-review + codex) on this PR surfaced: - renderer's advertised CJS entry was broken: it keeps @openmaic/dsl external and imports a runtime enum from it, but dsl is ESM-only (no `require` condition), so `require('@openmaic/renderer')` would throw ERR_PACKAGE_PATH_NOT_EXPORTED. Make renderer ESM-only: drop the `.cjs` rollup output, `main` now points at the ESM build, and the `require` conditions are removed from `exports`. (importer is unaffected — it bundles dsl, so its CJS build still works.) - prepublishOnly re-ran the test suite during `pnpm -r publish`, so a flaky test after dsl had already published gave a non-atomic partial release. Reduce prepublishOnly to a build-only guard (never ship stale/empty dist) and move the real test/typecheck gate into the workflow, before any publish. - document that an @openmaic/* tag publishes the whole family via `pnpm -r` (pnpm skips already-published versions); the tag is a release marker, not a per-package gate. Verified: dsl + renderer + importer build; renderer emits ESM only (0 .cjs), all exports entries resolve; `npm pack` ships dist + README + LICENSE with no src leak; frozen-lockfile passes. Refs #778. * style: reflow @openmaic/dsl type imports past print-width after rename The @maic -> @openmaic rename lengthened two single-line type imports past prettier's 100-col width; prettier --check flagged them. Pure formatting. Refs #778. * docs(importer): mark @openmaic/importer browser-only (cr-loop accepted limitation) codex cross-review flagged that the published @openmaic/importer throws `XMLHttpRequest is not a constructor` when loaded in a pure Node process — its rollup build is browser-targeted (`nodeResolve({browser:true})` + a browser pdf.js build). The app only consumes it client-side ('use client'), so this is by design. Document it as an accepted limitation: prominent browser-only note in the README and a `browser` field in the manifest. Refs #778. |
||
|
wyuc
|
1d1ce80e04 |
feat(maic-agent): editor-agent line (v0) — Pro-mode "Edit with AI" (#777)
Promote the editor-agent line to main: read-then-act editor agent (read_scene_content, regenerate_scene, regenerate_scene_actions, edit_interactive_html), a routable maic-agent model stage, the agent sidebar UI, interactive-scene runtime-error capture, and the timeline/script editor. Cross-reviewed and smoke-verified; see #777. |
||
|
wyucandClaude Opus 4.8
|
13b42206f6 |
fix(generation): don't regenerate a deleted slide on a finished deck (#769)
Deleting a slide in Pro/Edit mode removed it from `scenes` but left its entry in the persisted `outlines` plan. On the next classroom mount the resume-generation pass treats the now-missing `order` as a pending outline and regenerates the slide, appending it as the last page. Rather than prune outlines by `order` (which is rebalanced by Pro-mode insert/duplicate/reorder, so it is not a stable key and would mis-target or break delete-undo), gate resume on a persisted `generationComplete` flag: - The flag is recorded (alongside outlines, in the stageOutlines record) when generation finishes — both generateRemaining completion paths, the retry-to-completion path (markGenerationCompleteIfDone), and the already-materialized-at-mount path — and restored on load. It is written only after a verified scene flush, so it never outruns the deck's scenes. - A finished deck is frozen for editing: loadFromStorage no longer derives pending placeholders from orphaned outlines, the classroom resume effect skips generateRemaining, media resume is filtered to materialized outlines, and isCourseComplete honours the flag — so a deleted slide neither regenerates nor leaves a phantom pending page. - Decks generated before the flag existed self-heal on load (every outline has a scene ⇒ inferred complete). deleteScene additionally records completion when deleting from a deck that is complete right now, so the "Course complete" end page and resume-suppression survive even when the flag was never persisted (e.g. edited without a reload). `outlines` are never pruned, so order-rebalancing and delete-undo are unaffected. Genuinely interrupted generations (not complete) still resume. Fixes #768 Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com> |
||
|
wyucandwyuc
|
267965d6ed |
feat(generation): infer concise courseTitle from outlines for readable course names (#756)
* feat(generation): infer concise courseTitle from outlines for readable course names
Generated courses were named after the raw user prompt (truncated to 500 chars
client-side, or 50 chars / first outline title server-side), which is long and
unreadable. The outline-generation step already understands the full course
semantics, so have it emit a concise `courseTitle` alongside `languageDirective`
and `outlines`, and use it as the stage name.
- Prompts (requirements-to-outlines, interactive-outlines): top-level JSON gains
a required `courseTitle` (≤30 chars, in the teaching language, noun phrase).
- Streaming route: extract `courseTitle` head-bound like `languageDirective`,
emit a `courseTitle` SSE event, and include it in the `done` event.
- Non-streaming outline-generator: parse and return optional `courseTitle`
(defensive trim + 120-char cap).
- Naming: stage.name = courseTitle || outlines[0]?.title || requirement fallback
(client preview path backfills stage.name after outlines resolve; server
classroom-generation uses it directly).
- Session: persist `courseTitle` on GenerationSessionState for reload safety.
- Tests: add courseTitle parsing coverage; e2e mock done event carries it.
Fully optional in the pipeline — every consumer falls back to the previous
behavior when the field is absent, so legacy/LLM-skipped cases are unaffected.
* fix(generation): make courseTitle propagation robust across all outline paths
Address review feedback on the courseTitle threading.
Prompts — align every outline-prompt schema statement to the 3-key shape.
The user-prompt "final reminder" / top-level shape blocks for
requirements-to-outlines, interactive-outlines, and task-engine-outlines still
demanded the old two-key {languageDirective, outlines} object; as the last
instruction the model reads, that caused it to omit courseTitle and the stage
name to fall back to the raw requirement. Task Engine prompts now emit
courseTitle as well.
Streaming route — normalize the streamed courseTitle (trim, ignore
whitespace-only, cap length) to match the non-streaming parser, and add a
full-buffer fallback so a title emitted after the outlines array or beyond the
head-scan window is still recovered before the done event.
Client — reset latched languageDirective/title on outline retry so a
succeeding attempt that omits them falls back instead of inheriting a failed
attempt's stale value; also carry courseTitle in the stream-end resolve path
that fires when the stream closes without an explicit done event.
courseTitle remains optional on every consumer path.
---------
Co-authored-by: wyuc <zdq1204@gmail.com>
|
||
|
wyucandClaude Opus 4.8
|
25cf58d19f |
feat(model): optional per-stage LLM model routing (#745) (#747)
* feat(model): optional per-stage LLM model routing (#745) Add a config-only stage -> model map consulted during model resolution, falling back to DEFAULT_MODEL when unset (zero behavior change unless opted in). - lib/server/model-routes.ts: parse/validate/cache MODEL_ROUTES env JSON; LLM_STAGES registry of routable stages; getStageModel(stage). - resolveModel: resolution order x-model > stage route > DEFAULT_MODEL > builtin; thread optional `stage` through resolveModelFromHeaders/FromRequest. - Wire each route's resolveModel call site to its stage; classroom-generation resolves generate-classroom and web-search-query-rewrite independently. - .env.example: document MODEL_ROUTES; unit tests for both modules. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * fix(model): address cross-review findings for per-stage routing (#745) - classroom-generation: resolve the web-search-query-rewrite model lazily, only when explicitly routed and inside the web-search branch, with try/catch fallback to the classroom model. Fixes a regression where a misconfigured optional route (keyless/invalid provider) threw and aborted ALL classroom generation even with web search disabled; also avoids wasted resolution on the common no-search path and preserves classroom-model inheritance when unrouted. - resolve-model: type the stage param as LlmStage across resolveModel/ resolveModelFromHeaders/resolveModelFromRequest so a mistyped stage literal is a compile error instead of silently falling through to DEFAULT_MODEL. - model-routes: simplify the cache to a process singleton (matching provider-config), dropping the raw-string-keyed cache and redundant STAGE_SET. - .env.example: clarify x-model precedence and that a route to an unconfigured provider fails at request time (no startup validation), like a bad DEFAULT_MODEL. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * fix(model): stage route takes precedence over client x-model (#745) Cross-review (codex P2 + Claude tracer) found that the browser UI always sends its saved model as x-model, so the previous x-model > stage order made MODEL_ROUTES inert for normal UI traffic on exactly the heavy stages the RFC targets (scene-content, quiz, pbl-chat, chat). Flip resolution to stage route > x-model > DEFAULT_MODEL: a configured route is the operator's deliberate per-stage choice and wins, while unrouted stages still honor the client's x-model. Update docs and tests accordingly. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * style(model): prettier-format changed files for #745 CI Run the repo formatter (pnpm check / prettier) on the changed routes, the new model-routes module, and the new tests so `prettier . --check` (CI) passes; also correct a stale precedence comment in the resolve-model test. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * fix(model): isolate routed model from client connection params (#745) When a stage route overrides the client's x-model and points at a different provider, the client-sent apiKey/baseUrl/providerType (for the client's model) must not bleed onto the routed provider — otherwise a routed Anthropic model would be constructed with the client's OpenAI providerType/key and fail. A routed model now resolves its connection params purely from server config, as if no x-model was sent. Unrouted stages still honor the client params. Adds tests covering both the drop (routed) and keep (unrouted) cases. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(model): correct graceful-degradation comment in classroom (#745) resolveModel does not throw on a missing key; clarify that a keyless route surfaces later in callLLM (degraded by the outer try/catch), while only a resolution-time failure is caught at the rewrite re-resolve. Comment-only. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * feat(model): route scene-content per scene type (#745) Extend MODEL_ROUTES with composite keys scene-content:<type> for the four core scene types (slide/quiz/interactive/pbl). getStageModel now resolves composite `a:b` stages most-specific-first, trimming `:` segments, so scene-content:<type> falls back to the base scene-content route, then x-model, then DEFAULT_MODEL. The scene-content route derives its stage from outline.type. Backward compatible: with no composite key configured, behavior is unchanged. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * feat(model): per-stage thinking effort + drop hardcoded gpt fallback (#745) MODEL_ROUTES values may now be {model, effort} (not just a model string) to pin the thinking effort per stage. Arbitration mirrors model routing: a routed stage with effort set wins over the client's thinking (effort "none" disables); routed with no effort uses the model's default and drops client thinking; unrouted stages keep the client thinking. resolveModel is the single arbiter — body thinking is threaded in via resolveModelFromRequest. classroom-generation passes the resolved thinking into its callLLM calls so generate-classroom and web-search-query-rewrite honor route effort too. Also: removed the hardcoded `|| 'gpt-5.4-mini'` fallback in resolveModel — if no model resolves (no route / x-model / DEFAULT_MODEL) it now throws instead of silently picking a vendor default. model-routes: getStageRoute returns {model, effort?}; getStageModel delegates to it. Docs + tests updated. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * refactor(model): route value carries full ThinkingConfig (#745) Per-stage routing accepts the full unified ThinkingConfig in the route value ({model, thinking:{mode,effort,level,enabled,budgetTokens,excludeReasoningOutput}}) instead of just an effort string. The route's thinking is passed through resolveModel and normalized per the model's capability by callLLM, so budgetTokens (qwen), level (Gemini), enabled/mode, etc. all work. (qwen3.7-plus/max thinking capability now comes from upstream #753, so this no longer registers them.) Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * fix(model): honor routed thinking for chat-adapter; doc fixes (#745) Cross-review (codex P2 + Claude) found chat-adapter was the only routable stage that ignored the route's thinking: it took the routed model but built its own thinkingConfig from the request body, so operator-pinned thinking never applied and client thinking still leaked onto a routed model. Now pass the client thinking into resolveModel and use the resolved thinkingConfig (route-pinned for a routed stage, client otherwise; defaults to disabled for low-latency chat). Also: clarify DEFAULT_MODEL is now required for server-side stages (the hardcoded gpt-5.4-mini fallback was removed → resolveModel throws if nothing resolves), and fix a stale {model,effort}→{model,thinking} doc comment. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com> |
||
|
wyucandClaude Opus 4.8
|
4dce1f7414 |
refactor(types): import the slide DSL directly from @maic/dsl; drop the shim (#738)
#707 left lib/types/slides.ts as a thin `export * from '@maic/dsl'` shim so the ~100 existing `@/lib/types/slides` import sites kept working unchanged. Point them at `@maic/dsl` directly and delete the shim, so the app consumes the package contract with no indirection and the legacy module path is gone. Pure module-specifier migration — every symbol imported from `@/lib/types/slides` was already a re-export of the same `@maic/dsl` symbol, so behavior is identical. Also redirects the two `lib/types` relative importers (`./slides`), the e2e fixture's relative import, and the dynamic `import('../types/slides').Slide` type alias in stage-storage; merges the now same-module import pair in slide-edit-elements. Verified: tsc --noEmit clean, eslint clean, vitest 804/804, next build OK, e2e (Playwright) 19/19. Part of #720 (Phase 1). Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com> |
||
|
|
1c858960b4 |
refactor(app): consume @maic/dsl + @maic/renderer in the app (slide types + read-only thumbnails) (#707)
* refactor(app): consume @maic/dsl + @maic/renderer for slide types and read-only thumbnails
The app extracted @maic/{dsl,renderer,importer} (#668) but kept rendering its
own duplicated copies. Start dogfooding the packages back into the app:
1. lib/types/slides.ts is now a thin re-export of @maic/dsl. The hand-kept
copy was the seed for @maic/dsl, so consuming the package stops the three
former copies (app / renderer / importer) from drifting again. All
`from '@/lib/types/slides'` import sites keep working unchanged.
(ShapePathFormulasKeys / ElementTypes move from local `const enum` to the
package's regular `enum` — required across a package boundary; runtime
value usage is identical.)
2. Read-only slide thumbnails now render via @maic/renderer's SlideCanvas
through a small SlideThumbnail wrapper, replacing the in-app ThumbnailSlide
element renderers at the three read-only surfaces: playback scene sidebar,
scene thumbnail content (editor rail), and home recent-course cards. The
wrapper preserves the legacy thumbnail video treatment (muted play-badged
<video>, placeholder filtering) via the renderer's renderVideo slot. The
full-size editing canvas is untouched (renderer v1 is read-only).
Verified: tsc --noEmit clean, next build OK, e2e 19/19 green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
* fix(thumbnail): resolve gen_* media placeholders so retries reflect in thumbnails
SlideThumbnail rendered slides through @maic/renderer with only a renderVideo
slot, so image (and video) `gen_img_*`/`gen_vid_*` placeholders were painted
as-is. @maic/renderer is a pure package and does not know about this app's
async media-generation store; the store also never mutates the slide's
`element.src` (it keeps the placeholder ref and tracks the generated objectUrl
in a task). Net effect inside a classroom: a thumbnail showed the broken
placeholder on first paint and — the reported bug — did NOT update when a
failed image's retry finally succeeded, because nothing re-resolved the src.
Add `useResolvedSlide`: resolve each placeholder element's `src` to its task
`objectUrl` (video `poster` too) against the media-generation store,
reactively. Unresolved (pending/failed) placeholders are blanked so the
renderer paints nothing instead of a broken-media icon. Off-classroom (no
stage context, e.g. home recent-course cards) renders raw, matching the legacy
thumbnail's behavior. Mirrors the app's existing `useResolvedImageSrc` resolver
that the full-size canvas already uses — restoring the reactive sync the
in-app ThumbnailElement had before the @maic/renderer migration.
The store subscription is a primitive signature of just this slide's
placeholder resolutions (src → done objectUrl), so unrelated task churn —
other slides' media generating or retrying — neither re-renders the thumbnail
nor invalidates the resolved-slide memo.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* fix(thumbnail): key video tasks by mediaRef; use a visible signature delimiter
Review follow-ups (#707):
- Videos' media-generation tasks are keyed by mediaRef (with gen_vid_* src as
the legacy fallback), not by element src — resolve them through
getVideoMediaRefForElement, matching BaseVideoElement's subscription, so a
mediaRef-keyed video thumbnail updates when generation or a retry completes.
A mediaRef-keyed video that already carries a real playable src keeps it
while the task is unresolved; only placeholder srcs are blanked.
- The resolution-signature delimiter was a literal NUL byte, which made git
treat the source file as binary (degrading diffs and review tooling). Use a
visible '|' delimiter instead.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: 杨慎 <117187635+cosarah@users.noreply.github.com>
|