LangChain
Framework for building applications with LLMs through composable tools and chains.
MCPCLI
About LangChain
Framework for building applications with LLMs through composable tools and chains. Explore how LangChain integrates with the agentic data stack ecosystem and supports autonomous data operations.
Key Features
- Open-source agent engineering platform with pre-built agent architecture and 1000+ integrations
- Modular building blocks: chains, agents, memory, tools, and indexes for complex AI workflows
- Retrieval-Augmented Generation (RAG) with vector database integrations for grounding LLM responses
- Dynamic agent reasoning: analyze context, decide actions, invoke tools, and iterate
- LangServe for exposing chains and agents as REST APIs with a single command
- LangSmith integration for observability, evaluation, tracing, and deployment
- LangGraph for building stateful, multi-actor applications with cycles and branching
- Vendor-agnostic: swap models, tools, and databases without rewriting application code
Agent Integration
MCP Server
langchain-ai/langchain-mcp-adaptersExternal Links
LangChain MCP Adapters
Official library converting MCP tools into LangChain/LangGraph-compatible tools, supports stdio and HTTP
MCP Documentation
Official MCP integration guide in LangChain docs
LangGraph Agent Framework
Graph-based agent orchestration framework recommended for production agents with MCP
Python API Reference
Comprehensive API reference for all langchain-* packages
LangChain Deep Agents
Agent harness with planning, filesystem backend, and subagent spawning for complex tasks