langchain 0.2.8 → 0.2.10

package/README.md

@@ -2,14 +2,13 @@
 
 ⚡ Building applications with LLMs through composability ⚡
 
-[![CI](https://github.com/langchain-ai/langchainjs/actions/workflows/ci.yml/badge.svg)](https://github.com/langchain-ai/langchainjs/actions/workflows/ci.yml) ![npm](https://img.shields.io/npm/dm/langchain) [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT) [![Twitter](https://img.shields.io/twitter/url/https/twitter.com/langchainai.svg?style=social&label=Follow%20%40LangChainAI)](https://twitter.com/langchainai) [![](https://dcbadge.vercel.app/api/server/6adMQxSpJS?compact=true&style=flat)](https://discord.gg/6adMQxSpJS) [![Open in Dev Containers](https://img.shields.io/static/v1?label=Dev%20Containers&message=Open&color=blue&logo=visualstudiocode)](https://vscode.dev/redirect?url=vscode://ms-vscode-remote.remote-containers/cloneInVolume?url=https://github.com/langchain-ai/langchainjs)
+[![CI](https://github.com/langchain-ai/langchainjs/actions/workflows/ci.yml/badge.svg)](https://github.com/langchain-ai/langchainjs/actions/workflows/ci.yml) ![npm](https://img.shields.io/npm/dm/langchain) [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT) [![Twitter](https://img.shields.io/twitter/url/https/twitter.com/langchainai.svg?style=social&label=Follow%20%40LangChainAI)](https://twitter.com/langchainai) [![Open in Dev Containers](https://img.shields.io/static/v1?label=Dev%20Containers&message=Open&color=blue&logo=visualstudiocode)](https://vscode.dev/redirect?url=vscode://ms-vscode-remote.remote-containers/cloneInVolume?url=https://github.com/langchain-ai/langchainjs)
 [<img src="https://github.com/codespaces/badge.svg" title="Open in Github Codespace" width="150" height="20">](https://codespaces.new/langchain-ai/langchainjs)
 
 Looking for the Python version? Check out [LangChain](https://github.com/langchain-ai/langchain).
 
 To help you ship LangChain apps to production faster, check out [LangSmith](https://smith.langchain.com).
 [LangSmith](https://smith.langchain.com) is a unified developer platform for building, testing, and monitoring LLM applications.
-Fill out [this form](https://airtable.com/appwQzlErAS2qiP0L/shrGtGaVBVAz7NcV2) to get off the waitlist or speak with our sales team.
 
 ## ⚡️ Quick Install
 
@@ -17,10 +16,6 @@ You can use npm, yarn, or pnpm to install LangChain.js
 
 `npm install -S langchain` or `yarn add langchain` or `pnpm add langchain`
 
-```typescript
-import { ChatOpenAI } from "langchain/chat_models/openai";
-```
-
 ## 🌐 Supported Environments
 
 LangChain is written in TypeScript and can be used in:
@@ -39,19 +34,20 @@ LangChain is written in TypeScript and can be used in:
 - **Reason**: rely on a language model to reason (about how to answer based on provided context, what actions to take, etc.)
 
 This framework consists of several parts.
-- **LangChain Libraries**: The Python and JavaScript libraries. Contains interfaces and integrations for a myriad of components, a basic runtime for combining these components into chains and agents, and off-the-shelf implementations of chains and agents.
-- **[LangChain Templates](https://github.com/langchain-ai/langchain/tree/master/templates)**: (currently Python-only) A collection of easily deployable reference architectures for a wide variety of tasks.
-- **[LangServe](https://github.com/langchain-ai/langserve)**: (currently Python-only) A library for deploying LangChain chains as a REST API.
-- **[LangSmith](https://smith.langchain.com)**: A developer platform that lets you debug, test, evaluate, and monitor chains built on any LLM framework and seamlessly integrates with LangChain.
+- **Open-source libraries**: Build your applications using LangChain's open-source [building blocks](https://js.langchain.com/v0.2/docs/concepts#langchain-expression-language), [components](https://js.langchain.com/v0.2/docs/concepts), and [third-party integrations](https://js.langchain.com/v0.2/docs/integrations/platforms/).
+Use [LangGraph.js](https://js.langchain.com/v0.2/docs/concepts/#langgraphjs) to build stateful agents with first-class streaming and human-in-the-loop support.
+- **Productionization**: Use [LangSmith](https://docs.smith.langchain.com/) to inspect, monitor and evaluate your chains, so that you can continuously optimize and deploy with confidence.
+- **Deployment**: Turn your LangGraph applications into production-ready APIs and Assistants with [LangGraph Cloud](https://langchain-ai.github.io/langgraph/cloud/) (currently Python-only).
 
 The LangChain libraries themselves are made up of several different packages.
 - **[`@langchain/core`](https://github.com/langchain-ai/langchainjs/blob/main/langchain-core)**: Base abstractions and LangChain Expression Language.
 - **[`@langchain/community`](https://github.com/langchain-ai/langchainjs/blob/main/libs/langchain-community)**: Third party integrations.
 - **[`langchain`](https://github.com/langchain-ai/langchainjs/blob/main/langchain)**: Chains, agents, and retrieval strategies that make up an application's cognitive architecture.
+- **[LangGraph.js](https://langchain-ai.github.io/langgraphjs/)**: A library for building robust and stateful multi-actor applications with LLMs by modeling steps as edges and nodes in a graph. Integrates smoothly with LangChain, but can be used without it.
 
 Integrations may also be split into their own compatible packages.
 
-![LangChain Stack](https://github.com/langchain-ai/langchainjs/blob/main/docs/core_docs/static/img/langchain_stack_feb_2024.webp)
+![LangChain Stack](https://github.com/langchain-ai/langchainjs/blob/main/docs/core_docs/static/svg/langchain_stack_062024.svg)
 
 This library aims to assist in the development of those types of applications. Common examples of these applications include:
 
@@ -86,15 +82,15 @@ Data Augmented Generation involves specific types of chains that first interact
 
 **🤖 Agents:**
 
-Agents involve an LLM making decisions about which Actions to take, taking that Action, seeing an Observation, and repeating that until done. LangChain provides a standard interface for agents, a selection of agents to choose from, and examples of end-to-end agents.
+Agents allow an LLM autonomy over how a task is accomplished. Agents make decisions about which Actions to take, then take that Action, observe the result, and repeat until the task is complete. LangChain provides a [standard interface for agents](https://js.langchain.com/v0.2/docs/concepts/#agents), along with [LangGraph.js](https://github.com/langchain-ai/langgraphjs/) for building custom agents.
 
 ## 📖 Documentation
 
-Please see [here](https://js.langchain.com/v0.2/) for full documentation, which includes:
+Please see [here](https://js.langchain.com) for full documentation, which includes:
 
 - [Getting started](https://js.langchain.com/v0.2/docs/introduction): installation, setting up the environment, simple examples
-- [Tutorials](https://js.langchain.com/v0.2/docs/tutorials/): interactive guides and walkthroughs of common use cases/tasks.
-- [Use case](https://js.langchain.com/v0.2/docs/how_to/) walkthroughs and best practices for every component of the LangChain library.
+- Overview of the [interfaces](https://js.langchain.com/v0.2/docs/how_to/lcel_cheatsheet/), [modules](https://js.langchain.com/v0.2/docs/concepts) and [integrations](https://js.langchain.com/v0.2/docs/integrations/platforms/)
+- [Tutorial](https://js.langchain.com/v0.2/docs/tutorials/) walkthroughs
 - [Reference](https://api.js.langchain.com): full API docs
 
 ## 💁 Contributing
@@ -108,4 +104,3 @@ Please report any security issues or concerns following our [security guidelines
 ## 🖇️ Relationship with Python LangChain
 
 This is built to integrate as seamlessly as possible with the [LangChain Python package](https://github.com/langchain-ai/langchain). Specifically, this means all objects (prompts, LLMs, chains, etc) are designed in a way where they can be serialized and shared between languages.
-

package/dist/agents/executor.cjs

@@ -478,6 +478,9 @@ class AgentExecutor extends base_js_1.BaseChain {
                     observation = tool
                         ? await tool.invoke(action.toolInput, (0, runnables_1.patchConfig)(config, { callbacks: runManager?.getChild() }))
                         : `${action.tool} is not a valid tool, try another one.`;
+                    if (typeof observation !== "string") {
+                        throw new Error("Received unsupported non-string response from tool call.");
+                    }
                 }
                 catch (e) {
                     // eslint-disable-next-line no-instanceof/no-instanceof
@@ -573,6 +576,9 @@ class AgentExecutor extends base_js_1.BaseChain {
                 const tool = nameToolMap[agentAction.tool];
                 try {
                     observation = await tool.call(agentAction.toolInput, runManager?.getChild());
+                    if (typeof observation !== "string") {
+                        throw new Error("Received unsupported non-string response from tool call.");
+                    }
                 }
                 catch (e) {
                     // eslint-disable-next-line no-instanceof/no-instanceof

package/dist/agents/executor.js

@@ -473,6 +473,9 @@ export class AgentExecutor extends BaseChain {
                     observation = tool
                         ? await tool.invoke(action.toolInput, patchConfig(config, { callbacks: runManager?.getChild() }))
                         : `${action.tool} is not a valid tool, try another one.`;
+                    if (typeof observation !== "string") {
+                        throw new Error("Received unsupported non-string response from tool call.");
+                    }
                 }
                 catch (e) {
                     // eslint-disable-next-line no-instanceof/no-instanceof
@@ -568,6 +571,9 @@ export class AgentExecutor extends BaseChain {
                 const tool = nameToolMap[agentAction.tool];
                 try {
                     observation = await tool.call(agentAction.toolInput, runManager?.getChild());
+                    if (typeof observation !== "string") {
+                        throw new Error("Received unsupported non-string response from tool call.");
+                    }
                 }
                 catch (e) {
                     // eslint-disable-next-line no-instanceof/no-instanceof

package/dist/document_loaders/tests/csv-blob.test.js

@@ -28,6 +28,7 @@ test("Test CSV loader from blob", async () => {
     expect(docs.length).toBe(2);
     expect(docs[0]).toMatchInlineSnapshot(`
     Document {
+      "id": undefined,
       "metadata": {
         "blobType": "text/csv",
         "line": 1,
@@ -39,6 +40,7 @@ test("Test CSV loader from blob", async () => {
   `);
     expect(docs[1]).toMatchInlineSnapshot(`
     Document {
+      "id": undefined,
       "metadata": {
         "blobType": "text/csv",
         "line": 2,

package/dist/document_loaders/tests/json-blob.test.js

@@ -24,6 +24,7 @@ test("Test JSON loader from blob", async () => {
     expect(docs.length).toBe(2);
     expect(docs[0]).toMatchInlineSnapshot(`
     Document {
+      "id": undefined,
       "metadata": {
         "blobType": "application/json",
         "line": 1,
@@ -34,6 +35,7 @@ test("Test JSON loader from blob", async () => {
   `);
     expect(docs[1]).toMatchInlineSnapshot(`
     Document {
+      "id": undefined,
       "metadata": {
         "blobType": "application/json",
         "line": 2,
@@ -66,6 +68,7 @@ test("Test JSON loader from blob", async () => {
     expect(docs.length).toBe(10);
     expect(docs[0]).toMatchInlineSnapshot(`
     Document {
+      "id": undefined,
       "metadata": {
         "blobType": "application/json",
         "line": 1,
@@ -76,6 +79,7 @@ test("Test JSON loader from blob", async () => {
   `);
     expect(docs[1]).toMatchInlineSnapshot(`
     Document {
+      "id": undefined,
       "metadata": {
         "blobType": "application/json",
         "line": 2,

package/dist/document_loaders/tests/jsonl-blob.test.js

@@ -23,6 +23,7 @@ test("Test JSONL loader from blob", async () => {
     expect(docs.length).toBe(2);
     expect(docs[0]).toMatchInlineSnapshot(`
     Document {
+      "id": undefined,
       "metadata": {
         "blobType": "application/jsonl+json",
         "line": 1,
@@ -33,6 +34,7 @@ test("Test JSONL loader from blob", async () => {
   `);
     expect(docs[1]).toMatchInlineSnapshot(`
     Document {
+      "id": undefined,
       "metadata": {
         "blobType": "application/jsonl+json",
         "line": 2,

package/dist/retrievers/parent_document.cjs

@@ -160,6 +160,17 @@ class ParentDocumentRetriever extends multi_vector_js_1.MultiVectorRetriever {
         parentDocs.push(...retrievedDocs);
         return parentDocs.slice(0, this.parentK);
     }
+    async _storeDocuments(parentDoc, childDocs, addToDocstore) {
+        if (this.childDocumentRetriever) {
+            await this.childDocumentRetriever.addDocuments(childDocs);
+        }
+        else {
+            await this.vectorstore.addDocuments(childDocs);
+        }
+        if (addToDocstore) {
+            await this.docstore.mset(Object.entries(parentDoc));
+        }
+    }
     /**
      * Adds documents to the docstore and vectorstores.
      * If a retriever is provided, it will be used to add documents instead of the vectorstore.
@@ -192,8 +203,6 @@ class ParentDocumentRetriever extends multi_vector_js_1.MultiVectorRetriever {
         if (parentDocs.length !== parentDocIds.length) {
             throw new Error(`Got uneven list of documents and ids.\nIf "ids" is provided, should be same length as "documents".`);
         }
-        const embeddedDocs = [];
-        const fullDocs = {};
         for (let i = 0; i < parentDocs.length; i += 1) {
             const parentDoc = parentDocs[i];
             const parentDocId = parentDocIds[i];
@@ -202,17 +211,7 @@ class ParentDocumentRetriever extends multi_vector_js_1.MultiVectorRetriever {
                 pageContent: subDoc.pageContent,
                 metadata: { ...subDoc.metadata, [this.idKey]: parentDocId },
             }));
-            embeddedDocs.push(...taggedSubDocs);
-            fullDocs[parentDocId] = parentDoc;
-        }
-        if (this.childDocumentRetriever) {
-            await this.childDocumentRetriever.addDocuments(embeddedDocs);
-        }
-        else {
-            await this.vectorstore.addDocuments(embeddedDocs);
-        }
-        if (addToDocstore) {
-            await this.docstore.mset(Object.entries(fullDocs));
+            await this._storeDocuments({ [parentDocId]: parentDoc }, taggedSubDocs, addToDocstore);
         }
     }
 }

package/dist/retrievers/parent_document.d.ts

@@ -63,6 +63,7 @@ export declare class ParentDocumentRetriever extends MultiVectorRetriever {
     documentCompressorFilteringFn?: ParentDocumentRetrieverFields["documentCompressorFilteringFn"];
     constructor(fields: ParentDocumentRetrieverFields);
     _getRelevantDocuments(query: string): Promise<Document[]>;
+    _storeDocuments(parentDoc: Record<string, Document>, childDocs: Document[], addToDocstore: boolean): Promise<void>;
     /**
      * Adds documents to the docstore and vectorstores.
      * If a retriever is provided, it will be used to add documents instead of the vectorstore.

package/dist/retrievers/parent_document.js

@@ -134,6 +134,17 @@ export class ParentDocumentRetriever extends MultiVectorRetriever {
         parentDocs.push(...retrievedDocs);
         return parentDocs.slice(0, this.parentK);
     }
+    async _storeDocuments(parentDoc, childDocs, addToDocstore) {
+        if (this.childDocumentRetriever) {
+            await this.childDocumentRetriever.addDocuments(childDocs);
+        }
+        else {
+            await this.vectorstore.addDocuments(childDocs);
+        }
+        if (addToDocstore) {
+            await this.docstore.mset(Object.entries(parentDoc));
+        }
+    }
     /**
      * Adds documents to the docstore and vectorstores.
      * If a retriever is provided, it will be used to add documents instead of the vectorstore.
@@ -166,8 +177,6 @@ export class ParentDocumentRetriever extends MultiVectorRetriever {
         if (parentDocs.length !== parentDocIds.length) {
             throw new Error(`Got uneven list of documents and ids.\nIf "ids" is provided, should be same length as "documents".`);
         }
-        const embeddedDocs = [];
-        const fullDocs = {};
         for (let i = 0; i < parentDocs.length; i += 1) {
             const parentDoc = parentDocs[i];
             const parentDocId = parentDocIds[i];
@@ -176,17 +185,7 @@ export class ParentDocumentRetriever extends MultiVectorRetriever {
                 pageContent: subDoc.pageContent,
                 metadata: { ...subDoc.metadata, [this.idKey]: parentDocId },
             }));
-            embeddedDocs.push(...taggedSubDocs);
-            fullDocs[parentDocId] = parentDoc;
-        }
-        if (this.childDocumentRetriever) {
-            await this.childDocumentRetriever.addDocuments(embeddedDocs);
-        }
-        else {
-            await this.vectorstore.addDocuments(embeddedDocs);
-        }
-        if (addToDocstore) {
-            await this.docstore.mset(Object.entries(fullDocs));
+            await this._storeDocuments({ [parentDocId]: parentDoc }, taggedSubDocs, addToDocstore);
         }
     }
 }

package/dist/retrievers/tests/ensemble_retriever.int.test.js

@@ -11,7 +11,7 @@ test("Should work with a question input", async () => {
         "Cars are made out of plastic",
         "mitochondria is the powerhouse of the cell",
         "mitochondria is made of lipids",
-    ], [{ id: 1 }, { id: 2 }, { id: 3 }, { id: 4 }, { id: 5 }], new CohereEmbeddings());
+    ], [{ id: 1 }, { id: 2 }, { id: 3 }, { id: 4 }, { id: 5 }], new CohereEmbeddings({ model: "embed-english-v3.0" }));
     const retriever = new EnsembleRetriever({
         retrievers: [vectorstore.asRetriever()],
     });
@@ -28,7 +28,7 @@ test("Should work with multiple retriever", async () => {
         "Cars are made out of plastic",
         "mitochondria is the powerhouse of the cell",
         "mitochondria is made of lipids",
-    ], [{ id: 1 }, { id: 2 }, { id: 3 }, { id: 4 }, { id: 5 }], new CohereEmbeddings());
+    ], [{ id: 1 }, { id: 2 }, { id: 3 }, { id: 4 }, { id: 5 }], new CohereEmbeddings({ model: "embed-english-v3.0" }));
     const vectorstore2 = await MemoryVectorStore.fromTexts([
         "Buildings are made out of brick",
         "Buildings are made out of wood",
@@ -37,7 +37,7 @@ test("Should work with multiple retriever", async () => {
         "Cars are made out of plastic",
         "mitochondria is the powerhouse of the cell",
         "mitochondria is made of lipids",
-    ], [{ id: 6 }, { id: 7 }, { id: 8 }, { id: 9 }, { id: 10 }], new CohereEmbeddings());
+    ], [{ id: 6 }, { id: 7 }, { id: 8 }, { id: 9 }, { id: 10 }], new CohereEmbeddings({ model: "embed-english-v3.0" }));
     const retriever = new EnsembleRetriever({
         retrievers: [vectorstore.asRetriever(), vectorstore2.asRetriever()],
     });
@@ -54,7 +54,7 @@ test("Should work with weights", async () => {
         "Cars are made out of plastic",
         "mitochondria is the powerhouse of the cell",
         "mitochondria is made of lipids",
-    ], [{ id: 1 }, { id: 2 }, { id: 3 }, { id: 4 }, { id: 5 }], new CohereEmbeddings());
+    ], [{ id: 1 }, { id: 2 }, { id: 3 }, { id: 4 }, { id: 5 }], new CohereEmbeddings({ model: "embed-english-v3.0" }));
     const vectorstore2 = await MemoryVectorStore.fromTexts([
         "Buildings are made out of brick",
         "Buildings are made out of wood",
@@ -63,7 +63,7 @@ test("Should work with weights", async () => {
         "Cars are made out of plastic",
         "mitochondria is the powerhouse of the cell",
         "mitochondria is made of lipids",
-    ], [{ id: 6 }, { id: 7 }, { id: 8 }, { id: 9 }, { id: 10 }], new CohereEmbeddings());
+    ], [{ id: 6 }, { id: 7 }, { id: 8 }, { id: 9 }, { id: 10 }], new CohereEmbeddings({ model: "embed-english-v3.0" }));
     const retriever = new EnsembleRetriever({
         retrievers: [vectorstore.asRetriever(), vectorstore2.asRetriever()],
         weights: [0.5, 0.9],

package/dist/retrievers/tests/multi_query.int.test.js

@@ -12,7 +12,7 @@ test("Should work with a question input", async () => {
         "Cars are made out of plastic",
         "mitochondria is the powerhouse of the cell",
         "mitochondria is made of lipids",
-    ], [{ id: 1 }, { id: 2 }, { id: 3 }, { id: 4 }, { id: 5 }], new CohereEmbeddings());
+    ], [{ id: 1 }, { id: 2 }, { id: 3 }, { id: 4 }, { id: 5 }], new CohereEmbeddings({ model: "embed-english-v3.0" }));
     const model = new ChatOpenAI({});
     const retriever = MultiQueryRetriever.fromLLM({
         llm: model,
@@ -32,7 +32,7 @@ test("Should work with a keyword", async () => {
         "Cars are made out of plastic",
         "mitochondria is the powerhouse of the cell",
         "mitochondria is made of lipids",
-    ], [{ id: 1 }, { id: 2 }, { id: 3 }, { id: 4 }, { id: 5 }], new CohereEmbeddings());
+    ], [{ id: 1 }, { id: 2 }, { id: 3 }, { id: 4 }, { id: 5 }], new CohereEmbeddings({ model: "embed-english-v3.0" }));
     const model = new ChatOpenAI({});
     const retriever = MultiQueryRetriever.fromLLM({
         llm: model,

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "langchain",
-  "version": "0.2.8",
+  "version": "0.2.10",
   "description": "Typescript bindings for langchain",
   "type": "module",
   "engines": {
@@ -888,7 +888,7 @@
     }
   },
   "dependencies": {
-    "@langchain/core": ">=0.2.9 <0.3.0",
+    "@langchain/core": ">=0.2.11 <0.3.0",
     "@langchain/openai": ">=0.1.0 <0.3.0",
     "@langchain/textsplitters": "~0.0.0",
     "binary-extensions": "^2.2.0",
@@ -900,7 +900,7 @@
     "ml-distance": "^4.0.0",
     "openapi-types": "^12.1.3",
     "p-retry": "4",
-    "uuid": "^9.0.0",
+    "uuid": "^10.0.0",
     "yaml": "^2.2.1",
     "zod": "^3.22.4",
     "zod-to-json-schema": "^3.22.3"