This project provides a Streamlit interface for LangGraph-powered ReAct agents that leverage the Model Context Protocol (MCP) for tool integration. It allows users to dynamically configure, deploy, and interact with AI agents capable of accessing various data sources and APIs, offering real-time streaming responses and conversation history.

How It Works

The project utilizes LangGraph for agent orchestration and integrates with the Model Context Protocol (MCP). MCP facilitates agent-tool interaction through a client-server architecture: MCP Hosts (like LangGraph) use MCP Clients to communicate with MCP Servers, which expose specific functionalities. This design allows agents to seamlessly access external data and APIs through a standardized protocol.

Quick Start & Requirements

• Docker: cd dockers && docker compose -f docker-compose.yaml up -d (AMD64) or docker compose -f docker-compose-mac.yaml up -d (ARM64).
• Source: git clone ... && cd langgraph-mcp-agents && uv venv && uv pip install -r requirements.txt && source .venv/bin/activate.
• Prerequisites: API keys for Anthropic, OpenAI, and/or LangSmith (optional). Docker Desktop for Docker installation.
• Access: http://localhost:8585
• Docs: MCP-HandsOn-KOR.ipynb

Highlighted Details

• Streamlit interface for dynamic tool management (add, remove, configure via Smithery JSON).
• Supports real-time streaming responses and conversation history.
• Integrates with Anthropic and OpenAI models, with optional LangSmith tracing.
• Provides a hands-on Jupyter notebook tutorial for deeper MCP and LangGraph integration insights.

Maintenance & Community

No specific contributor or community links (Discord/Slack) are mentioned in the README.

Licensing & Compatibility

MIT License. Permissive for commercial use and closed-source linking.

Limitations & Caveats

The project relies on external API keys for LLM access and LangSmith tracing. Specific MCP server implementations are not detailed within this README, requiring reference to external resources like Smithery.