This PR was opened by the [Changesets release](https://github.com/changesets/action) GitHub action. When you're ready to do a release, you can merge this and the packages will be published to npm automatically. If you're not ready to do a release yet, that's fine, whenever you add more changesets to main, this PR will be updated. # Releases ## @langchain/langgraph-checkpoint-validation@1.1.0 ### Minor Changes - [#2503](https://github.com/langchain-ai/langgraphjs/pull/2503) [`6e850fb`](https://github.com/langchain-ai/langgraphjs/commit/6e850fbabc842bb495352fbd6ef424c67ab0948f) Thanks [@jkennedyvz](https://github.com/jkennedyvz)! - fix(langgraph-checkpoint-validation): align node engine with vitest ## @langchain/langgraph-api@1.3.0 ### Minor Changes - [#2505](https://github.com/langchain-ai/langgraphjs/pull/2505) [`cad31b4`](https://github.com/langchain-ai/langgraphjs/commit/cad31b42f001a87fcdf57c4c084c655c8762b6a5) Thanks [@christian-bromann](https://github.com/christian-bromann)! - Consolidate the protocol session's channel inference, channel-set validation, and namespace prefix matching onto the shared `@langchain/langgraph/stream` helpers instead of maintaining local copies. This removes duplicated logic across `session/index.mts`, `session/namespace.mts`, `session/internal-types.mts`, and `service.mts`. This also aligns SSE event-sink filtering (`matchesSinkFilter`) with the WebSocket subscription matcher: both now normalize dynamic namespace suffixes (e.g. a `["fetcher"]` namespace filter matches an event emitted under `["fetcher:<uuid>"]`). Previously the SSE path used a stricter exact-segment match. Because the session now imports `@langchain/langgraph/stream`, the `@langchain/langgraph` peer dependency floor is raised to `^1.3.6` (the first release that ships the `/stream` entrypoint). ### Patch Changes - Updated dependencies \[]: - @langchain/langgraph-ui@1.3.0 ## @langchain/langgraph-checkpoint-sqlite@1.0.2 ### Patch Changes - [#2504](https://github.com/langchain-ai/langgraphjs/pull/2504) [`e8a0940`](https://github.com/langchain-ai/langgraphjs/commit/e8a09409ac4a997012e78081160c91188ebe39fc) Thanks [@jackjin1997](https://github.com/jackjin1997)! - fix: `SqliteSaver.list({}, { filter })` now honors arbitrary metadata keys (e.g. `tenant_id`, `env`), matching the behavior of the MongoDB, Postgres, and Redis checkpointers. Previously only `source`, `step`, and `parents` were honored — any other key was silently dropped, returning unfiltered results. ## @langchain/langgraph-cli@1.3.0 ### Patch Changes - Updated dependencies \[[`cad31b4`](https://github.com/langchain-ai/langgraphjs/commit/cad31b42f001a87fcdf57c4c084c655c8762b6a5)]: - @langchain/langgraph-api@1.3.0 ## @langchain/langgraph@1.3.7 ### Patch Changes - [#2505](https://github.com/langchain-ai/langgraphjs/pull/2505) [`cad31b4`](https://github.com/langchain-ai/langgraphjs/commit/cad31b42f001a87fcdf57c4c084c655c8762b6a5) Thanks [@christian-bromann](https://github.com/christian-bromann)! - Add the `@langchain/langgraph/stream` entrypoint — a transport-agnostic backend toolkit for building custom servers on top of the v2 streaming protocol. Alongside the existing `StreamChannel` and `convertToProtocolEvent`, it exposes subscription primitives, typed against a minimal `MatchableEvent` shape so they work on both the core `ProtocolEvent` and the wire-level `Event` from `@langchain/protocol`: - `inferChannel(event)` — map an event to its subscription `Channel` (named `custom:<name>` channels included). - `matchesSubscription(event, definition)` — decide whether a buffered event should be delivered for a `SubscribeParams` filter, honoring channel, namespace prefix/depth, and an optional `since` replay cursor. - `isPrefixMatch(namespace, prefix)` / `normalizeNamespaceSegment(segment)` — namespace prefix matching with dynamic-suffix normalization (e.g. `fetcher:<uuid>` matches the `fetcher` prefix). - `SUPPORTED_CHANNELS` / `isSupportedChannel(value)` — the recognized channel set and a guard for validating subscription requests. - Updated dependencies \[[`cad31b4`](https://github.com/langchain-ai/langgraphjs/commit/cad31b42f001a87fcdf57c4c084c655c8762b6a5)]: - @langchain/langgraph-sdk@1.9.19 ## @langchain/langgraph-sdk@1.9.19 ### Patch Changes - [#2505](https://github.com/langchain-ai/langgraphjs/pull/2505) [`cad31b4`](https://github.com/langchain-ai/langgraphjs/commit/cad31b42f001a87fcdf57c4c084c655c8762b6a5) Thanks [@christian-bromann](https://github.com/christian-bromann)! - Deduplicate the client stream transports: the protocol transport now shares the SSE decoder and `IterableReadableStream` helpers with the legacy transport instead of carrying its own copies. Removes the redundant `transport/decoder.ts` and `transport/stream.ts` shims (and a dead `StreamPart` re-export), importing the shared utilities from `utils/sse.ts` directly. No public API or behavior change. ## @langchain/angular@1.0.19 ### Patch Changes - Updated dependencies \[[`cad31b4`](https://github.com/langchain-ai/langgraphjs/commit/cad31b42f001a87fcdf57c4c084c655c8762b6a5)]: - @langchain/langgraph-sdk@1.9.19 ## @langchain/react@1.0.19 ### Patch Changes - Updated dependencies \[[`cad31b4`](https://github.com/langchain-ai/langgraphjs/commit/cad31b42f001a87fcdf57c4c084c655c8762b6a5)]: - @langchain/langgraph-sdk@1.9.19 ## @langchain/svelte@1.0.19 ### Patch Changes - Updated dependencies \[[`cad31b4`](https://github.com/langchain-ai/langgraphjs/commit/cad31b42f001a87fcdf57c4c084c655c8762b6a5)]: - @langchain/langgraph-sdk@1.9.19 ## @langchain/vue@1.0.19 ### Patch Changes - Updated dependencies \[[`cad31b4`](https://github.com/langchain-ai/langgraphjs/commit/cad31b42f001a87fcdf57c4c084c655c8762b6a5)]: - @langchain/langgraph-sdk@1.9.19 ## @langchain/langgraph-ui@1.3.0 ## @example/ai-elements@0.1.34 ### Patch Changes - Updated dependencies \[[`cad31b4`](https://github.com/langchain-ai/langgraphjs/commit/cad31b42f001a87fcdf57c4c084c655c8762b6a5)]: - @langchain/langgraph@1.3.7 - @langchain/react@1.0.19 ## @examples/assistant-ui-claude@0.1.34 ### Patch Changes - Updated dependencies \[[`cad31b4`](https://github.com/langchain-ai/langgraphjs/commit/cad31b42f001a87fcdf57c4c084c655c8762b6a5)]: - @langchain/langgraph@1.3.7 - @langchain/react@1.0.19 ## @examples/ui-angular@0.0.44 ### Patch Changes - Updated dependencies \[[`cad31b4`](https://github.com/langchain-ai/langgraphjs/commit/cad31b42f001a87fcdf57c4c084c655c8762b6a5), [`cad31b4`](https://github.com/langchain-ai/langgraphjs/commit/cad31b42f001a87fcdf57c4c084c655c8762b6a5)]: - @langchain/langgraph@1.3.7 - @langchain/langgraph-sdk@1.9.19 - @langchain/angular@1.0.19 ## @examples/ui-multimodal@0.0.20 ### Patch Changes - Updated dependencies \[[`cad31b4`](https://github.com/langchain-ai/langgraphjs/commit/cad31b42f001a87fcdf57c4c084c655c8762b6a5)]: - @langchain/langgraph@1.3.7 - @langchain/react@1.0.19 ## @examples/ui-react@0.0.20 ### Patch Changes - Updated dependencies \[[`cad31b4`](https://github.com/langchain-ai/langgraphjs/commit/cad31b42f001a87fcdf57c4c084c655c8762b6a5), [`cad31b4`](https://github.com/langchain-ai/langgraphjs/commit/cad31b42f001a87fcdf57c4c084c655c8762b6a5)]: - @langchain/langgraph@1.3.7 - @langchain/langgraph-sdk@1.9.19 - @langchain/react@1.0.19 ## langgraph@1.0.39 ### Patch Changes - Updated dependencies \[[`cad31b4`](https://github.com/langchain-ai/langgraphjs/commit/cad31b42f001a87fcdf57c4c084c655c8762b6a5)]: - @langchain/langgraph@1.3.7 Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
@langchain/svelte
Svelte 5 SDK for Deep Agents, LangChain and LangGraph.
useStream binds a LangGraph agent into a Svelte 5 component. Reactive fields are exposed as getters on a stable handle (stream.messages, stream.isLoading, …) so templates and $derived expressions track updates automatically — no stores, no $ prefix, no destructuring.
Installation
npm install @langchain/svelte @langchain/core
Peer dependencies: svelte ^5.0.0, @langchain/core ^1.0.1
Quick start
<script lang="ts">
import { useStream } from "@langchain/svelte";
const stream = useStream({
assistantId: "agent",
apiUrl: "http://localhost:2024",
});
</script>
{#each stream.messages as msg (msg.id)}
<div>{msg.content}</div>
{/each}
<button
disabled={stream.isLoading}
onclick={() =>
stream.submit({ messages: [{ type: "human", content: "Hello!" }] })}
>
Send
</button>
Note: Access fields through the live
streamhandle. Destructuring (const { messages } = stream) freezes the values at that moment — usestream.messagesin templates instead.
Highlights
- v2-native streaming protocol. Session-based transport with automatic re-attach on remount; no more
reconnectOnMount/joinStreamdance. - Always-on root projections.
values,messages,toolCalls, andinterruptsare reactive at the root with zero extra subscription cost. - Selector composables for scoped data. Per-subagent / per-subgraph messages, tool calls, and media stream only when a component actually mounts the matching composable, and release on unmount.
- Discriminated option bag. The hosted Agent Server path and the custom-adapter path are two arms of a single typed union — mixing them is a compile-time error.
- Reactive
threadId. PassthreadId: () => activeto drive in-place thread swaps without remounting. - Agent-brand type inference.
useStream<typeof agent>()unwraps state, tool calls, and subagent state maps from an agent brand. - Multimodal media streams. Built-in assembly for audio, images, video, and files — plus opinionated playback helpers.
- Headless tools. Register local tool implementations that auto-resolve server-emitted tool-call interrupts without a round-trip through the UI.
Documentation
In-depth guides live in docs/:
useStream— options, return shape, reactivethreadId- Selector composables (
useMessages,useToolCalls,useValues, …) - Interrupts,
respond(),stop(),hydrationPromise - Submission queue
- Stream context (
provideStream/getStream) - Headless tools
- Custom transport (
AgentServerAdapter,HttpAgentServerAdapter) - Media (images, audio, video, files) & playback helpers
- Type safety
Migrating from v0
@langchain/svelte v1 targets the v2 streaming protocol. The useStream import stays the same, but the option bag, return shape, and how you subscribe to scoped data all change. Most chat apps migrate in well under an hour — the full guide with line-by-line diffs lives in docs/v1-migration.md.
Playground
For full end-to-end examples, see the LangChain UI Playground.
License
MIT