chore(deps): update ai sdk - #157
Conversation
dawsontoth
left a comment
dawsontoth
left a comment
There was a problem hiding this comment.
Approving. Verified locally (Node 24.19.0, npm 11.17.0), worktree at 6f93b5d:
npm ci exit 0
npm run build exit 0 (tsup ESM + DTS)
npm run lint exit 0 (oxlint)
npm run format exit 0 (dprint check)
npm test 53 files, 345 tests passed
I also ran the check CI can't see
lifecycle/getModel.ts wraps every provider in aisdk() from @openai/agents-extensions/ai-sdk, which throws on an unexpected specificationVersion — and getModel.test.ts vi.mocks both @ai-sdk/anthropic and @openai/agents-extensions/ai-sdk, so the real gate never runs and all 345 tests pass regardless. An @ai-sdk/* bump that breaks every claude-* model at runtime would still show green. So I probed it directly:
anthropic specificationVersion= v4 aisdk() OK
openai specificationVersion= v4 aisdk() OK
google specificationVersion= v4 aisdk() OK
All three providers construct and pass through aisdk() unwrapped. The reason it's fine: this branch resolves @openai/agents-extensions@0.15.0, and extensions ≥0.14 declares '@ai-sdk/provider': '^2 || ^3 || ^4', so spec v4 is accepted. (It was extensions 0.13.4 — Only v2 and v3 are supported — that blocked #146.)
Versions moved
ai 7.0.83 → 7.0.92 with @ai-sdk/anthropic 4.0.49, google 4.0.63, openai 4.0.58, provider 4.0.10, provider-utils 5.0.36, gateway 4.0.74 — the whole set in lockstep, which is what keeps the spec version coherent.
Lockfile is intact
Checked explicitly, since the sibling #158 is broken by exactly this: 1649 keys / 28 react-native* entries, identical to main, and npm ci exits 0. No harper > alasql pruning on this branch.
All 5 checks green, MERGEABLE/CLEAN.
## [0.16.49](v0.16.48...v0.16.49) (2026-09-07) ### Dependency Updates * **deps:** update ai sdk ([#157](#157)) ([e1e6492](e1e6492))
|
🎉 This PR is included in version 0.16.49 🎉 The release is available on GitHub release Your semantic-release bot 📦🚀 |
1 participant
This PR contains the following updates:
4.0.44→4.0.494.0.56→4.0.634.0.50→4.0.587.0.83→7.0.92Release Notes
vercel/ai (@ai-sdk/anthropic)
v4.0.49Compare Source
Patch Changes
65397d7: fix(anthropic): recognize dated Google Vertex Claude 4 model IDs when selecting model capabilitiesv4.0.48Compare Source
Patch Changes
4d25a08: feat(anthropic): add fable 5.1 support6bcc0f8]v4.0.47Compare Source
Patch Changes
5190b67: feat(provider): extend the FilesV4 interface with optionalgetFileMetadata,downloadFile(streaming), anddeleteFileoperations, plusabortSignal/headerscall options and a{ type: 'stream' }upload data variant; upload results now exposebyteSize,createdAt, andexpiresAt(also surfaced by the coreuploadFile()helper, which now forwardsabortSignal/headers); addpostMultipartStreamToApi(streaming multipart uploads with deterministic part ordering and failure-path stream teardown),deleteFromApi, andcreateBinaryStreamResponseHandlerto provider-utils5190b67]v4.0.46Compare Source
Patch Changes
aa45741: fix(provider/anthropic): preserve native message batch request counts in provider metadata and support the full language-model option surface in batch requestsaa45741]v4.0.45Compare Source
Patch Changes
90192f1]vercel/ai (@ai-sdk/google)
v4.0.63Compare Source
Patch Changes
18ad19c: feat(google): support agentic video processing in the Interactions API048ce06: feat(batch): surface the uploaded input file on the batch start result (providerMetadata.<provider>.inputFileId/inputFileExpiresAt) and accept aninputFileExpiresAfterprovider option on the OpenAI and xAI batch input file upload622fa7f: feat(google): add gemini-3.8-flash modelv4.0.62Compare Source
Patch Changes
6bcc0f8]v4.0.61Compare Source
Patch Changes
5190b67: feat(provider): extend the FilesV4 interface with optionalgetFileMetadata,downloadFile(streaming), anddeleteFileoperations, plusabortSignal/headerscall options and a{ type: 'stream' }upload data variant; upload results now exposebyteSize,createdAt, andexpiresAt(also surfaced by the coreuploadFile()helper, which now forwardsabortSignal/headers); addpostMultipartStreamToApi(streaming multipart uploads with deterministic part ordering and failure-path stream teardown),deleteFromApi, andcreateBinaryStreamResponseHandlerto provider-utils5190b67]v4.0.60Compare Source
Patch Changes
e07b577: feat: add tool calling support to batchv4.0.59Compare Source
Patch Changes
ca29e9b: fix(google): support video response formats in Gemini Interactionsv4.0.58Compare Source
Patch Changes
aa45741]v4.0.57Compare Source
Patch Changes
949ef93: fix(google): recognize Google Files URLs when using a baseUrl90192f1]vercel/ai (@ai-sdk/openai)
v4.0.58Compare Source
Patch Changes
b54e551: fix(openai): support documented Responses image generation options17d3436: feat(openai): add support for ultrafast service tierv4.0.57Compare Source
Patch Changes
048ce06: feat(batch): surface the uploaded input file on the batch start result (providerMetadata.<provider>.inputFileId/inputFileExpiresAt) and accept aninputFileExpiresAfterprovider option on the OpenAI and xAI batch input file upload7243530: fix(openai): preserve complete Responses API raw usage objectsv4.0.56Compare Source
Patch Changes
6bcc0f8]v4.0.55Compare Source
Patch Changes
3fc40db: feat(openai): implementgetFileMetadata,downloadFile(streaming), anddeleteFileon the OpenAI files interface, support streaming uploads via{ type: 'stream' }data, exposebyteSize/createdAt/expiresAton upload results, and threadabortSignal/headersthrough all file operations; blank and dot-segment file ids are rejected/encoded so they cannot retarget request paths5190b67]v4.0.54Compare Source
Patch Changes
f6fac50: fix(openai): handle null usage in responsese07b577: feat: add tool calling support to batchv4.0.53Compare Source
Patch Changes
7439d7a: Fix workflow deserialization for OpenAI Responses models.v4.0.52Compare Source
Patch Changes
1c68540: Preserve explicit prompt cache breakpoints on scalar Responses tool results.aa45741]v4.0.51Compare Source
Patch Changes
90192f1]vercel/ai (ai)
v7.0.92Compare Source
Patch Changes
a51cc94: fix(ai): preserve provider metadata from empty smooth stream deltasd1904d3: fix(ai): surface fallback errors for empty HTTP response bodies84e5a79: fix(ai): skipsmoothStreamdelays while the document is hiddena8e8ad0: fix(ai): expose call ID and abort reason in streamText onAbort callbacksa7e324b]v7.0.91Compare Source
Patch Changes
802af1e: Add configurable recovery for provider errors received afterstreamTextresponse streaming begins. Explicitly configuringstreamRetriesenables isolated retry attempts, including one bounded callback-directed recovery throughStreamTextOnErrorRetryCallbackwithstreamRetries: 0; recovered results and metadata reflect only the successful attempt, while the existingStreamTextOnErrorCallbackcontract and logging-only observer behavior remain compatible.5484f27]36eb7ee]622fa7f]v7.0.90Compare Source
Patch Changes
4d25a08]6bcc0f8]v7.0.89Compare Source
Patch Changes
5190b67: feat(provider): extend the FilesV4 interface with optionalgetFileMetadata,downloadFile(streaming), anddeleteFileoperations, plusabortSignal/headerscall options and a{ type: 'stream' }upload data variant; upload results now exposebyteSize,createdAt, andexpiresAt(also surfaced by the coreuploadFile()helper, which now forwardsabortSignal/headers); addpostMultipartStreamToApi(streaming multipart uploads with deterministic part ordering and failure-path stream teardown),deleteFromApi, andcreateBinaryStreamResponseHandlerto provider-utils5190b67]v7.0.88Compare Source
Patch Changes
8b6b756: fix(ai): prevent generateText from accepting responses that violate required tool choicese07b577: feat: add tool calling support to batchv7.0.87Compare Source
Patch Changes
850d863: fix(ai): preserve approval descriptors in UI message streamsv7.0.86Compare Source
Patch Changes
11109ae: feat(ai): support signed tool approvals in WorkflowAgent1329d5a]v7.0.85Compare Source
Patch Changes
55a9981: Ensure canonical hashes preserve undefined array element positions.dd32de2: fix(ai): sum Gateway image-generation costs across split requestsaa45741: fix(provider/anthropic): preserve native message batch request counts in provider metadata and support the full language-model option surface in batch requestscc29073: feat(ai): expose individual image generation callsd2507af]aa45741]v7.0.84Compare Source
Patch Changes
6669d69: Expose parsed structured output instreamTextend callbacks.a6463ca: fix(ai): allow tool approval secrets in ToolLoopAgent settings and prepareCalle604532: fix(ai): handle stateful and empty-match regular expressions in smoothStream805bbfc]90192f1]Configuration
📅 Schedule: (in timezone America/New_York)
🚦 Automerge: Enabled.
♻ Rebasing: Whenever PR is behind base branch, or you tick the rebase/retry checkbox.
👻 Immortal: This PR will be recreated if closed unmerged. Get config help if that's undesired.
This PR was generated by Mend Renovate. View the repository job log.