Skip to content

Commit 7a1ff0c

Browse files
hntrlhntrl
committed
chore(*): add changelogs
1 parent 29f83dc commit 7a1ff0c

File tree

37 files changed

+773
-6
lines changed

37 files changed

+773
-6
lines changed

‎libs/langchain-classic/CHANGELOG.md‎

Lines changed: 5 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,5 @@
1+
# @langchain/classic
2+
3+
## 1.0.0
4+
5+
This release updates the package for compatibility with LangChain v1.0. See the v1.0 [release notes](https://docs.langchain.com/oss/javascript/releases/langchain-v1) for details on what's new.

‎libs/langchain-community/CHANGELOG.md‎

Lines changed: 64 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,64 @@
1+
# @langchain/community
2+
3+
## 1.0.0
4+
5+
This release updates the package for compatibility with LangChain v1.0. See the v1.0 [release notes](https://docs.langchain.com/oss/javascript/releases/langchain-v1) for details on what's new.
6+
7+
## 0.3.57
8+
9+
### Patch Changes
10+
11+
- fd4691f: use `keyEncoder` instead of insecure cache key getter
12+
- Updated dependencies [fd4691f]
13+
- Updated dependencies [2f19cd5]
14+
- Updated dependencies [d38e9d6]
15+
- Updated dependencies [3c94076]
16+
- [email protected]
17+
- @langchain/openai@0.6.14
18+
19+
## 0.3.56
20+
21+
### Patch Changes
22+
23+
- 6da726f: feat(@langchain/community): add sagemaker endpoint - embedding support
24+
- 28dd44f: chore(couchbase): Deprecate CouchbaseVectorStore and create CouchbaseSearchVectorStore
25+
- 9adccfe: chore(@langchain/community): remove Dria retriever
26+
- 0a640ad: feat(langchain-community): add custom schema option for neon vector store
27+
- 940e087: fix(astra): replace deprecated 'namespace' param name
28+
- 8ac8edd: add support for advanced metadata filters in similarity search
29+
- e9d1136: create index aurora dsql
30+
- e0b48fd: fix(community): improve TogetherAI error handling for chat models
31+
- c10ea3e: allow any chars in delimited identifiers in hanavector
32+
- 9adccfe: chore(@langchain/community): remove Dria retriever
33+
- Updated dependencies [41bd944]
34+
- Updated dependencies [6019a7d]
35+
- Updated dependencies [54f542c]
36+
- Updated dependencies [707a768]
37+
- Updated dependencies [caf5579]
38+
- Updated dependencies [d60f40f]
39+
- @langchain/openai@0.6.12
40+
- [email protected]
41+
- @langchain/weaviate@0.2.3
42+
43+
## 0.3.55
44+
45+
### Patch Changes
46+
47+
- f201ab8: bump firebase-admin dependency (#8861)
48+
- f201ab8: use URL encoding for paths in github document laoder (#8860)
49+
50+
## 0.3.54
51+
52+
### Patch Changes
53+
54+
- 4a3f5af: Update import_constants.ts (#8747)
55+
- 8f9c617: postgres indexes getTime returning NaN due to missing alias
56+
- 4d26533: BM25Retriever: escape regex metacharacters in getTermFrequency to prevent crashes
57+
- 9f491d6: milvus: Fix upsert operations when autoId is false
58+
- 9649f20: add jira issue title to metadata for documents
59+
- 9543ba1: add personalAccessToken to jira loader
60+
- Updated dependencies [e0bd88c]
61+
- Updated dependencies [4a3f5af]
62+
- Updated dependencies [424360b]
63+
- [email protected]
64+
- @langchain/openai@0.6.10

‎libs/langchain-core/CHANGELOG.md‎

Lines changed: 95 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -1,5 +1,100 @@
11
# @langchain/core
22

3+
## 1.0.0
4+
5+
🎉 **LangChain v1.0** is here! This release provides a focused, production-ready foundation for building agents with significant improvements to the core abstractions and APIs. See the [release notes](https://docs.langchain.com/oss/javascript/releases/langchain-v1) for more details.
6+
7+
### ✨ Major Features
8+
9+
#### Standard content blocks
10+
11+
A new unified API for accessing modern LLM features across all providers:
12+
13+
- **New `contentBlocks` property**: Provides provider-agnostic access to reasoning traces, citations, built-in tools (web search, code interpreters, etc.), and other advanced LLM features
14+
- **Type-safe**: Full TypeScript support with type hints for all content block types
15+
- **Backward compatible**: Content blocks can be loaded lazily with no breaking changes to existing code
16+
17+
Example:
18+
19+
```typescript
20+
const response = await model.invoke([
21+
{ role: "user", content: "What is the weather in Tokyo?" },
22+
]);
23+
24+
// Access structured content blocks
25+
for (const block of response.contentBlocks) {
26+
if (block.type === "thinking") {
27+
console.log("Model reasoning:", block.thinking);
28+
} else if (block.type === "text") {
29+
console.log("Response:", block.text);
30+
}
31+
}
32+
```
33+
34+
For more information, see our guide on [content blocks](https://docs.langchain.com/oss/javascript/langchain/messages#content).
35+
36+
#### Enhanced Message API
37+
38+
Improvements to the core message types:
39+
40+
- **Structured content**: Better support for multimodal content with the new content blocks API
41+
- **Provider compatibility**: Consistent message format across all LLM providers
42+
- **Rich metadata**: Enhanced metadata support for tracking message provenance and transformations
43+
44+
### 🔧 Improvements
45+
46+
- **Better structured output generation**: Core abstractions for generating structured outputs in the main agent loop
47+
- **Improved type safety**: Enhanced TypeScript definitions across all core abstractions
48+
- **Performance optimizations**: Reduced overhead in message processing and runnable composition
49+
- **Better error handling**: More informative error messages and better error recovery
50+
51+
### 📦 Package Changes
52+
53+
The `@langchain/core` package remains focused on essential abstractions:
54+
55+
- Core message types and content blocks
56+
- Base runnable abstractions
57+
- Tool definitions and schemas
58+
- Middleware infrastructure
59+
- Callback system
60+
- Output parsers
61+
- Prompt templates
62+
63+
### 🔄 Migration Notes
64+
65+
**Backward Compatibility**: This release maintains backward compatibility with existing code. Content blocks are loaded lazily, so no changes are required to existing applications.
66+
67+
**New Features**: To take advantage of new features like content blocks and middleware:
68+
69+
1. Update to `@langchain/core@next`:
70+
71+
```bash
72+
npm install @langchain/[email protected]
73+
```
74+
75+
2. Use the new `contentBlocks` property to access rich content:
76+
77+
```typescript
78+
const response = await model.invoke(messages);
79+
console.log(response.contentBlocks); // New API
80+
console.log(response.content); // Legacy API still works
81+
```
82+
83+
3. For middleware and `createAgent`, install `langchain@next`:
84+
85+
```bash
86+
npm install [email protected] @langchain/[email protected]
87+
```
88+
89+
### 📚 Additional Resources
90+
91+
- [LangChain 1.0 Announcement](https://blog.langchain.com/langchain-langchain-1-0-alpha-releases/)
92+
- [Migration Guide](https://docs.langchain.com/oss/javascript/migrate/langchain-v1)
93+
- [Content Blocks Documentation](https://docs.langchain.com/oss/javascript/langchain/messages#content)
94+
- [Agents Documentation](https://docs.langchain.com/oss/javascript/langchain/agents)
95+
96+
---
97+
398
## 0.3.78
499

5100
### Patch Changes

‎libs/langchain-textsplitters/CHANGELOG.md‎

Lines changed: 5 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,5 @@
1+
# @langchain/textsplitters
2+
3+
## 1.0.0
4+
5+
This release updates the package for compatibility with LangChain v1.0. See the v1.0 [release notes](https://docs.langchain.com/oss/javascript/releases/langchain-v1) for details on what's new.

‎libs/langchain/CHANGELOG.md‎

Lines changed: 214 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -1,5 +1,219 @@
11
# langchain
22

3+
## 1.0.0
4+
5+
🎉 **LangChain v1.0** is here! This release provides a focused, production-ready foundation for building agents. We've streamlined the framework around three core improvements: **`createAgent`**, **standard content blocks**, and a **simplified package structure**. See the [release notes](https://docs.langchain.com/oss/javascript/releases/langchain-v1) for complete details.
6+
7+
### ✨ Major Features
8+
9+
#### `createAgent` - A new standard for building agents
10+
11+
`createAgent` is the new standard way to build agents in LangChain 1.0. It provides a simpler interface than `createReactAgent` from LangGraph while offering greater customization potential through middleware.
12+
13+
**Key features:**
14+
15+
- **Clean, intuitive API**: Build agents with minimal boilerplate
16+
- **Built on LangGraph**: Get persistence, streaming, human-in-the-loop, and time travel out of the box
17+
- **Middleware-first design**: Highly customizable through composable middleware
18+
- **Improved structured output**: Generate structured outputs in the main agent loop without additional LLM calls
19+
20+
Example:
21+
22+
```typescript
23+
import { createAgent } from "langchain";
24+
25+
const agent = createAgent({
26+
model: "anthropic:claude-sonnet-4-5-20250929",
27+
tools: [getWeather],
28+
systemPrompt: "You are a helpful assistant.",
29+
});
30+
31+
const result = await agent.invoke({
32+
messages: [{ role: "user", content: "What is the weather in Tokyo?" }],
33+
});
34+
35+
console.log(result.content);
36+
```
37+
38+
Under the hood, `createAgent` is built on the basic agent loop—calling a model using LangGraph, letting it choose tools to execute, and then finishing when it calls no more tools.
39+
40+
**Built on LangGraph features (work out of the box):**
41+
42+
- **Persistence**: Conversations automatically persist across sessions with built-in checkpointing
43+
- **Streaming**: Stream tokens, tool calls, and reasoning traces in real-time
44+
- **Human-in-the-loop**: Pause agent execution for human approval before sensitive actions
45+
- **Time travel**: Rewind conversations to any point and explore alternate paths
46+
47+
**Structured output improvements:**
48+
49+
- Generate structured outputs in the main loop instead of requiring an additional LLM call
50+
- Models can choose between calling tools or using provider-side structured output generation
51+
- Significant cost reduction by eliminating extra LLM calls
52+
53+
Example:
54+
55+
```typescript
56+
import { createAgent } from "langchain";
57+
import * as z from "zod";
58+
59+
const weatherSchema = z.object({
60+
temperature: z.number(),
61+
condition: z.string(),
62+
});
63+
64+
const agent = createAgent({
65+
model: "openai:gpt-4o-mini",
66+
tools: [getWeather],
67+
responseFormat: weatherSchema,
68+
});
69+
70+
const result = await agent.invoke({
71+
messages: [{ role: "user", content: "What is the weather in Tokyo?" }],
72+
});
73+
74+
console.log(result.structuredResponse);
75+
```
76+
77+
For more information, see [Agents documentation](https://docs.langchain.com/oss/javascript/langchain/agents).
78+
79+
#### Middleware
80+
81+
Middleware is what makes `createAgent` highly customizable, raising the ceiling for what you can build. Great agents require **context engineering**—getting the right information to the model at the right time. Middleware helps you control dynamic prompts, conversation summarization, selective tool access, state management, and guardrails through a composable abstraction.
82+
83+
**Prebuilt middleware** for common patterns:
84+
85+
```typescript
86+
import {
87+
createAgent,
88+
summarizationMiddleware,
89+
humanInTheLoopMiddleware,
90+
piiRedactionMiddleware,
91+
} from "langchain";
92+
93+
const agent = createAgent({
94+
model: "anthropic:claude-sonnet-4-5-20250929",
95+
tools: [readEmail, sendEmail],
96+
middleware: [
97+
piiRedactionMiddleware({ patterns: ["email", "phone", "ssn"] }),
98+
summarizationMiddleware({
99+
model: "anthropic:claude-sonnet-4-5-20250929",
100+
maxTokensBeforeSummary: 500,
101+
}),
102+
humanInTheLoopMiddleware({
103+
interruptOn: {
104+
sendEmail: {
105+
allowedDecisions: ["approve", "edit", "reject"],
106+
},
107+
},
108+
}),
109+
] as const,
110+
});
111+
```
112+
113+
**Custom middleware** with lifecycle hooks:
114+
115+
| Hook | When it runs | Use cases |
116+
| --------------- | ------------------------ | --------------------------------------- |
117+
| `beforeAgent` | Before calling the agent | Load memory, validate input |
118+
| `beforeModel` | Before each LLM call | Update prompts, trim messages |
119+
| `wrapModelCall` | Around each LLM call | Intercept and modify requests/responses |
120+
| `wrapToolCall` | Around each tool call | Intercept and modify tool execution |
121+
| `afterModel` | After each LLM response | Validate output, apply guardrails |
122+
| `afterAgent` | After agent completes | Save results, cleanup |
123+
124+
Example custom middleware:
125+
126+
```typescript
127+
import { createMiddleware } from "langchain";
128+
129+
const contextSchema = z.object({
130+
userExpertise: z.enum(["beginner", "expert"]).default("beginner"),
131+
});
132+
133+
const expertiseBasedToolMiddleware = createMiddleware({
134+
wrapModelCall: async (request, handler) => {
135+
const userLevel = request.runtime.context.userExpertise;
136+
if (userLevel === "expert") {
137+
const tools = [advancedSearch, dataAnalysis];
138+
return handler(request.replace("openai:gpt-5", tools));
139+
}
140+
const tools = [simpleSearch, basicCalculator];
141+
return handler(request.replace("openai:gpt-5-nano", tools));
142+
},
143+
});
144+
145+
const agent = createAgent({
146+
model: "anthropic:claude-sonnet-4-5-20250929",
147+
tools: [simpleSearch, advancedSearch, basicCalculator, dataAnalysis],
148+
middleware: [expertiseBasedToolMiddleware],
149+
contextSchema,
150+
});
151+
```
152+
153+
For more information, see the [complete middleware guide](https://docs.langchain.com/oss/javascript/langchain/middleware).
154+
155+
#### Simplified Package
156+
157+
LangChain v1 streamlines the `langchain` package namespace to focus on essential building blocks for agents. The package exposes only the most useful and relevant functionality (most re-exported from `@langchain/core` for convenience).
158+
159+
**What's in the core `langchain` package:**
160+
161+
- `createAgent` and agent-related utilities
162+
- Core message types and content blocks
163+
- Middleware infrastructure
164+
- Tool definitions and schemas
165+
- Prompt templates
166+
- Output parsers
167+
- Base runnable abstractions
168+
169+
### 🔄 Migration Notes
170+
171+
#### `@langchain/classic` for Legacy Functionality
172+
173+
Legacy functionality has moved to [`@langchain/classic`](https://www.npmjs.com/package/@langchain/classic) to keep the core package lean and focused.
174+
175+
**What's in `@langchain/classic`:**
176+
177+
- Legacy chains and chain implementations
178+
- The indexing API
179+
- [`@langchain/community`](https://www.npmjs.com/package/@langchain/community) exports
180+
- Other deprecated functionality
181+
182+
**To migrate legacy code:**
183+
184+
1. Install `@langchain/classic`:
185+
186+
```bash
187+
npm install @langchain/classic
188+
```
189+
190+
2. Update your imports:
191+
192+
```typescript
193+
import { ... } from "langchain"; // [!code --]
194+
import { ... } from "@langchain/classic"; // [!code ++]
195+
196+
import { ... } from "langchain/chains"; // [!code --]
197+
import { ... } from "@langchain/classic/chains"; // [!code ++]
198+
```
199+
200+
#### Upgrading to v1
201+
202+
Install the v1 packages:
203+
204+
```bash
205+
npm install [email protected] @langchain/[email protected]
206+
```
207+
208+
### 📚 Additional Resources
209+
210+
- [LangChain 1.0 Announcement](https://blog.langchain.com/langchain-langchain-1-0-alpha-releases/)
211+
- [Migration Guide](https://docs.langchain.com/oss/javascript/migrate/langchain-v1)
212+
- [Agents Documentation](https://docs.langchain.com/oss/javascript/langchain/agents)
213+
- [Middleware Guide](https://blog.langchain.com/agent-middleware/)
214+
215+
---
216+
3217
## 0.3.36
4218

5219
### Patch Changes

0 commit comments

Comments
 (0)