Repository navigation
fix(deps): update all non-major dependencies - #497
renovate-appwrite[bot] wants to merge 1 commit into
Conversation
renovate-appwrite
Bot
added
the
dependencies
🟡 Tier B · Needs changes before merging
This PR refreshes dependency versions and committed lockfiles across the language templates, including several SDKs and build tools. It also advances pinned GitHub Actions and the language versions used by CI jobs.
Fix with agent prompt### Issue 1
node/rag-with-langchain/package-lock.json:140
**LangChain OpenAI 1.8.0 drops the declared Node 20 support**
This lock resolves `@langchain/openai` 1.8.0, whose engine requires Node >=22, while `node/rag-with-langchain/package.json` still declares support for `>=20.0.0` (and its transitive `openai` 7.50.0 has the same Node floor). A deployment using Node 20, which the package manifest permits, is now outside the supported runtime range.
### Issue 2
node-typescript/sync-with-meilisearch/package-lock.json:52
**Meilisearch 0.62 renames the imported constructor**
The 0.62 client exports `Meilisearch`, not `MeiliSearch`: `npm run build` in this template now fails with TS2724. The same stale named import also prevents module loading in `node/sync-with-meilisearch/src/main.js` and `bun/sync-with-meilisearch/src/main.ts`.
### Issue 3
node-typescript/sync-with-qdrant/package-lock.json:25
**Qdrant 1.20 removes the `search` method**
The client removed `search` in 1.19 in favor of `query`, but `node-typescript/sync-with-qdrant/src/main.ts:38` still calls `client.search`. I ran `npm run build` in this template; it fails with TS2339, and the search endpoint cannot work with this client version.
### Issue 4
node-typescript/sync-with-qdrant/package-lock.json:25
**Qdrant 1.20 requires a newer Node runtime**
The locked client requires Node >=22 (and its `undici` dependency requires >=20.18.1), while this template still declares Node >=18.17 in `package.json` and documents Node 18 in its README. The advertised runtime is outside the supported range; raise the runtime requirement or keep a client compatible with Node 18.
### Issue 5
node-typescript/sync-with-meilisearch/package-lock.json:52
**Meilisearch 0.62 excludes the documented Node 18 runtime**
This client declares Node `^20.19.0 || >=22.12.0`, but the template still declares Node >=18.17 in `package.json` and documents Node 18 in its README. The new lock therefore no longer matches the template's advertised runtime; align the runtime requirement or use a compatible client.
### Issue 6
dotnet/starter/StarterTemplate.csproj:9
**Keep the .NET Appwrite range open to compatible patches**
The inclusive upper bound stops at 0.27.0, so compatible patch releases such as 0.27.1 are excluded. The repository dependency rule requires allowing compatible patch/minor updates while excluding the next breaking version.
### Issue 7
ruby/starter/Gemfile:3
**Keep the Ruby Appwrite range open to compatible patches**
`< 28.4.1` excludes 28.4.1 and all later 28.4.x patch releases; the same cap was added in `ruby/sync_with_meilisearch/Gemfile:3`. The repository dependency rule requires allowing compatible patch/minor updates while excluding the next breaking version.
### Issue 8
ruby/whatsapp-with-vonage/Gemfile:3-4
**Ruby json and dotenv ranges exclude the next compatible patches**
These upper bounds exclude the next patch releases (json 3.1.1 and dotenv 3.2.1) rather than the next breaking versions. That prevents compatible fixes from being selected and conflicts with the repository dependency rule.
---
For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.📂 Walkthrough · 9
⏳ Still open from earlier reviews · 7
Reviewed |
![appwrite[bot]](https://avatars.githubusercontent.com/in/325556?s=60&v=4){kind=small}
| "version": "0.40.0", | ||
| "resolved": "https://registry.npmjs.org/meilisearch/-/meilisearch-0.40.0.tgz", | ||
| "integrity": "sha512-BoRhQMr2mBFLEeCfsvPluksGb01IaOiWvV3Deu3iEY+yYJ4jdGTu+IQi5FCjKlNQ7/TMWSN2XUToSgvH1tj0BQ==", | ||
| "version": "0.62.0", |
There was a problem hiding this comment.
Meilisearch 0.62 renames the imported constructor
The 0.62 client exports Meilisearch, not MeiliSearch: npm run build in this template now fails with TS2724. The same stale named import also prevents module loading in node/sync-with-meilisearch/src/main.js and bun/sync-with-meilisearch/src/main.ts.
Prompt To Fix With AI
This is a comment left during a code review.
Path: node-typescript/sync-with-meilisearch/package-lock.json
Line: 52
Comment:
**Meilisearch 0.62 renames the imported constructor**
The 0.62 client exports `Meilisearch`, not `MeiliSearch`: `npm run build` in this template now fails with TS2724. The same stale named import also prevents module loading in `node/sync-with-meilisearch/src/main.js` and `bun/sync-with-meilisearch/src/main.ts`.
---
For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.🟠 Major · bug · Reply if this doesn't apply.
| "version": "1.18.0", | ||
| "resolved": "https://registry.npmjs.org/@qdrant/js-client-rest/-/js-client-rest-1.18.0.tgz", | ||
| "integrity": "sha512-/0dqX5uV9chC1DnYSnU4gNMrDqse/pt6hHg3Rqqpl5isH7xl1xSNvffjzBoxycDD79luWn7Ho6Rh/61sOs5DNw==", | ||
| "version": "1.20.0", |
There was a problem hiding this comment.
Qdrant 1.20 removes the search method
The client removed search in 1.19 in favor of query, but node-typescript/sync-with-qdrant/src/main.ts:38 still calls client.search. I ran npm run build in this template; it fails with TS2339, and the search endpoint cannot work with this client version.
Prompt To Fix With AI
This is a comment left during a code review.
Path: node-typescript/sync-with-qdrant/package-lock.json
Line: 25
Comment:
**Qdrant 1.20 removes the `search` method**
The client removed `search` in 1.19 in favor of `query`, but `node-typescript/sync-with-qdrant/src/main.ts:38` still calls `client.search`. I ran `npm run build` in this template; it fails with TS2339, and the search endpoint cannot work with this client version.
---
For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.🟠 Major · bug · Reply if this doesn't apply.
| "version": "1.18.0", | ||
| "resolved": "https://registry.npmjs.org/@qdrant/js-client-rest/-/js-client-rest-1.18.0.tgz", | ||
| "integrity": "sha512-/0dqX5uV9chC1DnYSnU4gNMrDqse/pt6hHg3Rqqpl5isH7xl1xSNvffjzBoxycDD79luWn7Ho6Rh/61sOs5DNw==", | ||
| "version": "1.20.0", |
There was a problem hiding this comment.
Qdrant 1.20 requires a newer Node runtime
The locked client requires Node >=22 (and its undici dependency requires >=20.18.1), while this template still declares Node >=18.17 in package.json and documents Node 18 in its README. The advertised runtime is outside the supported range; raise the runtime requirement or keep a client compatible with Node 18.
Prompt To Fix With AI
This is a comment left during a code review.
Path: node-typescript/sync-with-qdrant/package-lock.json
Line: 25
Comment:
**Qdrant 1.20 requires a newer Node runtime**
The locked client requires Node >=22 (and its `undici` dependency requires >=20.18.1), while this template still declares Node >=18.17 in `package.json` and documents Node 18 in its README. The advertised runtime is outside the supported range; raise the runtime requirement or keep a client compatible with Node 18.
---
For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.🟡 Minor · bug · Reply if this doesn't apply.
| "version": "0.40.0", | ||
| "resolved": "https://registry.npmjs.org/meilisearch/-/meilisearch-0.40.0.tgz", | ||
| "integrity": "sha512-BoRhQMr2mBFLEeCfsvPluksGb01IaOiWvV3Deu3iEY+yYJ4jdGTu+IQi5FCjKlNQ7/TMWSN2XUToSgvH1tj0BQ==", | ||
| "version": "0.62.0", |
There was a problem hiding this comment.
Meilisearch 0.62 excludes the documented Node 18 runtime
This client declares Node ^20.19.0 || >=22.12.0, but the template still declares Node >=18.17 in package.json and documents Node 18 in its README. The new lock therefore no longer matches the template's advertised runtime; align the runtime requirement or use a compatible client.
Prompt To Fix With AI
This is a comment left during a code review.
Path: node-typescript/sync-with-meilisearch/package-lock.json
Line: 52
Comment:
**Meilisearch 0.62 excludes the documented Node 18 runtime**
This client declares Node `^20.19.0 || >=22.12.0`, but the template still declares Node >=18.17 in `package.json` and documents Node 18 in its README. The new lock therefore no longer matches the template's advertised runtime; align the runtime requirement or use a compatible client.
---
For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.🟡 Minor · bug · Reply if this doesn't apply.
| </PropertyGroup> | ||
| <ItemGroup> | ||
| <PackageReference Include="Appwrite" Version="[0.21.2,0.22.0)" /> | ||
| <PackageReference Include="Appwrite" Version="[0.21.2,0.27.0]" /> |
There was a problem hiding this comment.
Keep the .NET Appwrite range open to compatible patches
The inclusive upper bound stops at 0.27.0, so compatible patch releases such as 0.27.1 are excluded. The repository dependency rule requires allowing compatible patch/minor updates while excluding the next breaking version.
Prompt To Fix With AI
This is a comment left during a code review.
Path: dotnet/starter/StarterTemplate.csproj
Line: 9
Comment:
**Keep the .NET Appwrite range open to compatible patches**
The inclusive upper bound stops at 0.27.0, so compatible patch releases such as 0.27.1 are excluded. The repository dependency rule requires allowing compatible patch/minor updates while excluding the next breaking version.
---
For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.🟡 Minor · maintainability · Reply if this doesn't apply.
| source "https://rubygems.org" | ||
|
|
||
| gem 'appwrite', '>= 27.0.0', '< 28.0.0' | ||
| gem 'appwrite', '>= 27.0.0', '< 28.4.1' |
There was a problem hiding this comment.
Keep the Ruby Appwrite range open to compatible patches
< 28.4.1 excludes 28.4.1 and all later 28.4.x patch releases; the same cap was added in ruby/sync_with_meilisearch/Gemfile:3. The repository dependency rule requires allowing compatible patch/minor updates while excluding the next breaking version.
Prompt To Fix With AI
This is a comment left during a code review.
Path: ruby/starter/Gemfile
Line: 3
Comment:
**Keep the Ruby Appwrite range open to compatible patches**
`< 28.4.1` excludes 28.4.1 and all later 28.4.x patch releases; the same cap was added in `ruby/sync_with_meilisearch/Gemfile:3`. The repository dependency rule requires allowing compatible patch/minor updates while excluding the next breaking version.
---
For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.🟡 Minor · maintainability · Reply if this doesn't apply.
| gem 'json', '>= 2.7.6', '< 3.1.1' | ||
| gem 'dotenv', '>= 2.8.1', '< 3.2.1' |
There was a problem hiding this comment.
Ruby json and dotenv ranges exclude the next compatible patches
These upper bounds exclude the next patch releases (json 3.1.1 and dotenv 3.2.1) rather than the next breaking versions. That prevents compatible fixes from being selected and conflicts with the repository dependency rule.
Prompt To Fix With AI
This is a comment left during a code review.
Path: ruby/whatsapp-with-vonage/Gemfile
Line: 3-4
Comment:
**Ruby json and dotenv ranges exclude the next compatible patches**
These upper bounds exclude the next patch releases (json 3.1.1 and dotenv 3.2.1) rather than the next breaking versions. That prevents compatible fixes from being selected and conflicts with the repository dependency rule.
---
For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.🟡 Minor · maintainability · Reply if this doesn't apply.
ea3f2fa to
a4a99a6
Compare
a4a99a6 to
d3895e5
Compare
| "version": "1.5.5", | ||
| "resolved": "https://registry.npmjs.org/@langchain/openai/-/openai-1.5.5.tgz", | ||
| "integrity": "sha512-wX7dwb9z4nf5FHXlIl/X2mk08pzonvRHCt1D4+s1zXLP0duYDC95j7dulPIQJ6fmhbyYQc9Ki8mEhY/D1lB8kw==", | ||
| "version": "1.8.0", |
There was a problem hiding this comment.
LangChain OpenAI 1.8.0 drops the declared Node 20 support
This lock resolves @langchain/openai 1.8.0, whose engine requires Node >=22, while node/rag-with-langchain/package.json still declares support for >=20.0.0 (and its transitive openai 7.50.0 has the same Node floor). A deployment using Node 20, which the package manifest permits, is now outside the supported runtime range.
Prompt To Fix With AI
This is a comment left during a code review.
Path: node/rag-with-langchain/package-lock.json
Line: 140
Comment:
**LangChain OpenAI 1.8.0 drops the declared Node 20 support**
This lock resolves `@langchain/openai` 1.8.0, whose engine requires Node >=22, while `node/rag-with-langchain/package.json` still declares support for `>=20.0.0` (and its transitive `openai` 7.50.0 has the same Node floor). A deployment using Node 20, which the package manifest permits, is now outside the supported runtime range.
---
For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.🟡 Minor · bug · Reply if this doesn't apply.
Note
This PR body was truncated due to platform limits.
This PR contains the following updates:
2.60.0→2.76.02.77.01.2.4→1.3.41.3.7(+2)1.5.5→1.8.01.8.3(+2)1.18.0→1.20.022.20.4→22.20.522.20.1→22.20.5[0.21.2,0.22.0)→[0.21.2,0.27.0]v6.0.0→v6.1.0v7.0.0→v7.1.0v7.0.0→v7.1.00.4.0→0.22.00.23.0'>= 27.0.0', '< 28.0.0'→'>= 27.0.0', '< 28.4.1'from: "13.2.2"→from: "13.3.0"v1.7.2→v1.8.12.54.0→2.61.12.62.0'>= 2.8.1', '< 3.0.0'→'>= 2.8.1', '< 3.2.1'0.45.3→0.45.4^0.25.12→^0.28.0v0.13.1→v0.17.01.26.10→1.27.31.27.4'>= 2.7.6', '< 3.0.0'→'>= 2.7.6', '< 3.1.1'1.5.4→1.6.11.6.4(+2)3.0.3→3.1.0>=0.30.0,<0.31.0→>=0.43.0,<0.44.0^0.40.0→^0.62.00.63.0^0.35.0→^0.62.00.63.06.21.0→6.22.029.0.0→29.4.029.1.0→29.4.010.2.1→10.3.110.5.0(+2)>=0.27.8,<0.28.0→>=0.28.1,<0.29.0^0.6.4→^0.22.03.9.6→3.9.103.10.03.9.9→3.9.103.10.03.11→3.15^0.29.4→^0.34.03.1→3.4v1.320.0→v1.336.0v1.338.0(+1)1.0.149→1.0.1542.37.2→2.40.01.53.3→1.54.15.4.5→5.9.32.10.1→2.15.00.11.2→0.22.14.0.0→4.1.012.2.1→12.3.0Warning
Some dependencies could not be looked up. Check the Dependency Dashboard for more information.
Release Notes
elevenlabs/elevenlabs-js (@elevenlabs/elevenlabs-js)
v2.76.0Compare Source
What's Changed
Full Changelog: elevenlabs/elevenlabs-js@v2.75.0...v2.76.0
v2.75.0Compare Source
What's Changed
Full Changelog: elevenlabs/elevenlabs-js@v2.74.0...v2.75.0
v2.74.0Compare Source
What's Changed
Full Changelog: elevenlabs/elevenlabs-js@v2.73.0...v2.74.0
v2.73.0Compare Source
What's Changed
New Contributors
Full Changelog: elevenlabs/elevenlabs-js@v2.72.0...v2.73.0
v2.72.0Compare Source
What's Changed
Full Changelog: elevenlabs/elevenlabs-js@v2.71.0...v2.72.0
v2.71.0Compare Source
What's Changed
Full Changelog: elevenlabs/elevenlabs-js@v2.70.0...v2.71.0
v2.70.0Compare Source
What's Changed
Full Changelog: elevenlabs/elevenlabs-js@v2.69.0...v2.70.0
v2.69.0Compare Source
What's Changed
New Contributors
Full Changelog: elevenlabs/elevenlabs-js@v2.68.0...v2.69.0
v2.68.0Compare Source
What's Changed
Full Changelog: elevenlabs/elevenlabs-js@v2.67.0...v2.68.0
v2.67.0Compare Source
What's Changed
Full Changelog: elevenlabs/elevenlabs-js@v2.66.0...v2.67.0
v2.66.0Compare Source
What's Changed
New Contributors
Full Changelog: elevenlabs/elevenlabs-js@v2.65.0...v2.66.0
v2.65.0Compare Source
What's Changed
Full Changelog: elevenlabs/elevenlabs-js@v2.64.0...v2.65.0
v2.64.0Compare Source
What's Changed
Full Changelog: elevenlabs/elevenlabs-js@v2.63.0...v2.64.0
v2.63.0Compare Source
What's Changed
Full Changelog: elevenlabs/elevenlabs-js@v2.62.0...v2.63.0
v2.62.0Compare Source
What's Changed
Full Changelog: elevenlabs/elevenlabs-js@v2.61.0...v2.62.0
v2.61.0Compare Source
What's Changed
Full Changelog: elevenlabs/elevenlabs-js@v2.60.0...v2.61.0
langchain-ai/langchainjs (@langchain/core)
v1.3.4Compare Source
Patch Changes
#12104
e006308Thanks @hntrl! - ExportomitPayloadfrom@langchain/core/tracers/tracer_langchain, aTracePolicyprocessor that records an empty payload while keeping the run.#12102
6d157aeThanks @hntrl! -LangChainTracernow still sends and ends a run when its trace policy'sprocessInputsorprocessOutputsthrows, with that payload dropped and the error set, instead of skipping the run or leaving it pending in LangSmith.v1.3.3Compare Source
Patch Changes
d92aba6Thanks @hntrl! - AddtracePolicytoRunnableSequence. ItsprocessInputs/processOutputsshape only whatLangChainTracersends to LangSmith for the sequence's own run. Batched sequence runs now each record their own output.v1.3.2Compare Source
Patch Changes
71fd01aThanks @thushanth-bengre-langchain! - feat(core): addaddNotificationMetadata,getNotificationMetadataand theNotificationMetadatatype for marking messages the agent harness adds.v1.3.1Compare Source
Patch Changes
#12026
178af31Thanks @eliornl! - fix(core): deliver callbacks once to a handler registered twice by merged configsMerging two runnable configs that both carry the same callback handler, as happens when an agent's
streamEvents()runs inside a LangGraph node with tracing on, registered that handler twice, so it received every callback twice: every streamed chunk arrived twice. The configured callback manager now keeps each handler once, in order, including a handler passed as a plain object of methods that merged configs wrapped twice.#12042
50343baThanks @sera-wi! - fix(core): include nodemetadatainGraph.toJSON()output when set, matching Python#12048
304d190Thanks @sera-wi! - fix(core): escapedrawMermaidnode metadata labels and leave out object values, so metadata with brackets or objects no longer breaks Mermaid parsingv1.3.0Compare Source
Minor Changes
#11986
2caf64dThanks @richardscarrott! - feat: stream tool-call args astool-call-deltaslicesTool-call and server-tool-call args now stream as
tool-call-deltaevents, which carry only the new slice of the args, instead ofblock-deltasnapshots of the args so far. A large tool call now costs about the same bytes on the wire as the same length of text._streamResponseChunks), the OpenAI-compatible converter (OpenAI Completions, Groq, Mistral, IBM, OpenRouter, Perplexity, DeepSeek, Fireworks, Together AI and xAI Completions), OpenAI and xAI Responses, Anthropic (client and server tool calls) and Bedrock Converse. Google and Ollama still send one complete snapshot per call.id,nameandindexgo oncontent-block-start, which carries no args. Anidornamethat arrives later reaches consumers oncontent-block-finish.block-deltafields.argsmid-stream, or write agent stream transformers that do, readtool-call-deltainstead.await stream,.toolCallsand the assembled message are unchanged.tool-call-delta. You need@langchain/langgraph-sdk1.12.6 or later;@langchain/react,@langchain/vue,@langchain/svelteor@langchain/angular1.2.6 or later; andlangchain-core1.6.13 or later for Python clients.@langchain/corebelow 1.3.0. That also keeps you on@langchain/openai,@langchain/anthropic,@langchain/awsand@langchain/xaiversions that send snapshots.v1.2.20Compare Source
Patch Changes
d71c542Thanks @richardscarrott! - feat(core): addmidConversationSystemMessagesandmidConversationToolstoModelProfile. They say whether a chat model sends a system message after the first turn in place, and whether it can add a tool through a message rather than the request's tool list. Either way the prompt prefix before it stays unchanged.v1.2.19Compare Source
Patch Changes
#11982
5af4684Thanks @richardscarrott! - feat(core): readtool-call-deltainChatModelStreamChatModelStreamappendstool-call-deltaslices to a tool call's args, and still replaces them onblock-deltaUnknown delta typev1.2.18Compare Source
Patch Changes
#11956
93d62b0Thanks @hntrl! - Require langsmith >=0.11.0 to include the restored NullableHeaders type declarations.#11822
d3e6fe6Thanks @dqbd! - Add explicit agent address support toLangChainTracer, inherit project and address destinations together, and delegate default destination resolution to the LangSmith SDK.#11150
cea15cfThanks @mdnanocom! - fix(core): defer tool call chunk collapsing in AIMessageChunk to fix O(n²) streaming aggregationAIMessageChunkeagerly re-rancollapseToolCallChunksin its constructor on every streamed deltaconcat, re-parsing all accumulated tool args each time — O(n²) over a stream, which can block the event loop for seconds on long tool-call streams. Collapsing now runs lazily, once, on first read oftool_calls/invalid_tool_calls.v1.2.17Compare Source
Patch Changes
b1d9c7aThanks @KyleCodes! - Preserve runtime context in ambient runnable config across tool calls, nested runnables, and streaming.v1.2.16Compare Source
Patch Changes
bb50952Thanks @hntrl! - fix(core): expose branded class type for provider declarationsv1.2.14Compare Source
Patch Changes
bbed273Thanks @thushanth-bengre-langchain! - AddfileMimeTypestoModelProfileso models can advertise the MIME types they accept as generic file inputs. OpenAI Responses API models now report the file typesinput_filesupports;Chat Completions profiles are unchanged.
v1.2.13Compare Source
Patch Changes
a83dfb1Thanks @ccurme! - Abbreviate long tool-call IDs when formatting chat messages as strings, keeping original messages unchanged.v1.2.12Compare Source
Patch Changes
030a726Thanks @hntrl! - Preserve structured tool arguments in tracer run inputs.v1.2.11Compare Source
Patch Changes
#11603
fec9cd8Thanks @thushanth-bengre-langchain! - fix(core): build streamingllmOutput.tokenUsagefrom the fully-accumulated chunk instead of whichever individual chunk'susage_metadataarrived lastAffects both core streaming paths —
.stream()/.streamEvents()(_streamIterator) and.invoke()/.generate()when a streaming-preferring callback is attached (_generateWithCache'shasStreamingHandlerbranch). Previously,llmOutput.tokenUsagewas overwritten by each chunk in turn, so only the last chunk carryingusage_metadatawon — correct for providers that emit one cumulative total on a final chunk, but wrong for providers (e.g.@langchain/google,@langchain/anthropic) that emitusage_metadataas a per-chunk delta across multiple chunks, where the values must be summed.Note for provider authors: this assumes each streamed chunk's
usage_metadatais either a per-chunk delta or appears only on a single final chunk. A provider that instead repeats a cumulative total on every chunk will now see it summed (and inflated) inllmOutput.tokenUsage, matching the existing behavior of the correctly-workingmessage.usage_metadatafield.Also fixes
@langchain/google'sinvoke({streaming: true})path (no streaming-preferring callback attached), wherellmOutputwas never populated at all.#11590
ffebdc2Thanks @thushanth-bengre-langchain! - Fix OpenAI Responses API replay under Zero Data Retention when a response contains more than one reasoning item, for both v0 and v1. In v0, the default replay path now reusesresponse_metadata.outputdirectly, preserving every reasoning item'sid/encrypted_contentin original order. In v1,AIMessage.contentBlocks(outputVersion: "v1") is fixed the same way.additional_kwargs.reasoningis unchanged.v1.2.10Compare Source
Patch Changes
ff1248fThanks @afirstenberg! - fix(core): BaseMessage.text use contentBlocksv1.2.9Compare Source
Patch Changes
43e4396Thanks @thushanth-bengre-langchain! - Fix ChatVertexAI/ChatGoogle content blocks: includetool_callblocks frommessage.tool_callsand skip spurious emptytextblocks incontentBlocks.v1.2.8Compare Source
Patch Changes
#11369
d6ad973Thanks @hntrl! - fix(langchain): use unified endpoint for gateway#11342
3b0e4c4Thanks @thushanth-bengre-langchain! - feat(core): mark errors as retryable or not, and stop retrying the ones that aren'tRetry middleware retried every failure up to
maxRetries, including deterministic ones like a bad API key or an unknown model. Retries also nest, so a single such failure could cost dozens of API calls.@langchain/core/errorsaddsstampRetryable(error, retryable)andgetRetryable(error). Marking an error leaves its class and shape untouched, so a provider SDK error can be classified without breakinginstanceof.getRetryablereturnsundefinedfor errors nobody classified, and both are exported so tool authors can mark their own failures.modelRetryMiddlewareandtoolRetryMiddlewarenow respect the mark by default, and retries stop as soon as one is found rather than each layer spending its own budget. Aborted calls, context overflow, and oversized payloads are marked non-retryable out of the box.Models accept a per-call
maxRetriesso a surrounding retry loop can take over.Behavior change: errors marked non-retryable now fail on the first attempt. Unclassified errors — including any from third-party integrations or custom tools — retry exactly as before. Pass
retryOn: () => trueto restore the old default. A customonFailedAttemptreplaces the built-in handler and opts out of marking.v1.2.7Compare Source
Patch Changes
c068bbfThanks @hntrl! - fix(core,langchain): patch and release core, update peer dependenciesv1.2.6Compare Source
Patch Changes
#11344
f08e0c6Thanks @hntrl! - fix: apply [Symbol.hasInstance] method to all comparable properties using .isInstance()We have some internal schemas that rely on
z.instanceof(). This uses a strictinstanceofcheck which can conflict if there are multiple versions of core installed. This overrides the Symbol.hasInstance method to use the same logic as.isInstance()to compare objects at runtime.v1.2.5Compare Source
Patch Changes
#11305
e654022Thanks @jacoblee93! - Add LangSmith Gateway environment configuration to OpenAI, Anthropic, and Fireworks chat models.#11295
1a1b347Thanks @vladislav-nechakhin! - fix(core): pass the mustache escape override per render callqdrant/qdrant-js (@qdrant/js-client-rest)
v1.20.0Compare Source
Minor Changes
scoringon text index params, and atextquery forquery/queryBatch/queryGroups(field named byusing)substringmatch condition for keyword fieldsnestedoption on payload index params, to index every object of an array as its own entrymax,minandacoshformula expressionsturbo8andturbo16vector datatypes,bits8TurboQuant bit sizeid_tracker(point id tracker params) tocreateCollectionandrecreateCollection; collection info now reportscreated_atcodetokenizer for text indexes (source-code identifier splitting)v1.19.0Compare Source
Breaking Changes
Removed
search,searchBatch,searchPointGroups,recommend,recommendBatch,recommend_batch,recommendPointGroups,discoverPointsanddiscoverBatchPoints,following Qdrant dropping the corresponding endpoints from its OpenAPI spec. Use
query,queryBatchandqueryGroupsinstead:searchquerysearchBatchqueryBatchsearchPointGroupsqueryGroupsrecommendquerywith arecommendqueryrecommendBatchqueryBatchwithrecommendqueriesrecommendPointGroupsqueryGroupswith arecommendquerydiscoverPointsquerywith adiscoverorcontextquerydiscoverBatchPointsqueryBatchwithdiscoverorcontextqueriesThe request types that came with them are gone from
Schemasas well:SearchRequest,SearchRequestBatch,SearchGroupsRequest,RecommendRequest,RecommendRequestBatch,RecommendGroupsRequest,DiscoverRequest,DiscoverRequestBatch,RecommendExample,ContextExamplePair,NamedVector,NamedSparseVector,NamedVectorStructandUsingVector.Note that
queryand friends return{points: [...]}rather than a bare array.Minor Changes
getQuotasandupdateQuotasfor the cluster-wide quota APIpayload(payload storage params) andmetadataparameters tocreateCollectionandrecreateCollectionmax_disk_usage_percentmoved offstrict_mode_config— it is now a cluster-wide setting onupdateQuotas, alongsidemax_resident_memory_percentandrelease_margin_percentappwrite/sdk-for-dotnet (Appwrite)
v0.27.0Compare Source
v0.26.0Compare Source
dart-3.10andflutter-3.38toBuildRuntimeandRuntimeenumsbytesMaxandbytesUsedproperties toCollectionandTablemodelsRolesenum removed from Teams service;rolesparameter now acceptsIReadOnlyList<string>Backupsservicev0.25.0Compare Source
permissions: IReadOnlyList<BrowserPermission>).Outputenum has been removed; useImageFormatinstead.getQueueAuditssupport toHealthservice.DatabasesandTablesDBservices.v0.24.0Compare Source
v0.23.0Compare Source
VCSDeploymentTypeenum toVCSReferenceTypeCreateTemplateDeploymentmethod signature: replaceVersionparameter withType(TemplateReferenceType) andReferenceparametersGetScreenshotmethod toAvatarsserviceTheme,TimezoneandOutputenumsv0.22.0Compare Source
totalparameter to list queries allowing skipping counting rows in a table for improved performanceOperatorclass for atomic modification of rows via update, bulk update, upsert, and bulk upsert operationsCreateResendProviderandUpdateResendProvidermethods toMessagingserviceactions/setup-dotnet
v6.1.0Compare Source
What's Changed
Enhancements
Documentation
Dependency Updates