langchain-docs skill

---
name: langchain-docs
description: Local LangChain AI documentation reference. Use when asked about LangChain, LangGraph, agents, chains, prompts, memory, tools, retrieval, RAG, vector stores, document loaders, or building LLM applications.
---

# LangChain Documentation

LangChain is a framework for building LLM-powered applications. It provides components for chains, agents, RAG, memory, and tools. LangGraph extends this with stateful multi-agent orchestration.

## Navigation Guide

**LangChain:** `references/langchain/` - Core framework docs (51 files)
- Agents, chains, RAG, memory, tools, models, guardrails
- Key files: `overview.mdx`, `agents.mdx`, `knowledge-base.mdx`

**LangGraph:** `references/langgraph/` - Multi-agent orchestration (35 files)
- Graph API, memory, interrupts, durable execution
- Key files: `graph-api.mdx`, `memory.mdx`, `agentic-rag.mdx`

**Python SDK:** `references/python/` - Python API reference (1283 files)
- Complete API docs for langchain-core, langchain, langgraph

**JavaScript SDK:** `references/javascript/` - JS/TS API reference (299 files)
- Complete API docs for @langchain/core, langchain, langgraph

**Concepts:** `references/concepts/` - Foundational concepts (2 files)

**Deep Agents:** `references/deepagents/` - Advanced agent patterns (10 files)

## Key Entry Points

| Task | Start Here |
|------|------------|
| Getting started | `references/learn.mdx` |
| LangChain overview | `references/langchain/overview.mdx` |
| Build an agent | `references/langchain/agents.mdx` |
| RAG implementation | `references/langchain/knowledge-base.mdx` |
| LangGraph intro | `references/langgraph/graph-api.mdx` |
| Memory & state | `references/langgraph/memory.mdx` |
| Python API | `references/python/` |
| JavaScript API | `references/javascript/` |

## When to use

Use this skill when the user asks about:
- LangChain chains, agents, or components
- LangGraph multi-agent orchestration
- RAG (Retrieval Augmented Generation)
- Memory and conversation history
- Tools and tool calling
- Vector stores and embeddings
- Building LLM applications

## How to find information

1. **First**, read `references/STRUCTURE.md` to see all 1688 documentation files organized by directory
2. Use Navigation Guide to find the section
3. Check Key Entry Points for common tasks
4. For API details: `references/python/` or `references/javascript/`

**STRUCTURE.md contains a complete file listing - always check it first when searching for specific topics.**