# MCP-powered Ultimate AI Assistant A Streamlit application that provides a chat interface for interacting with MCP (Model Context Protocol) servers. This app allows you to configure multiple MCP servers and chat with them using natural language. Tech stack: - [mcp-use](https://github.com/mcp-use/mcp-use) to connect LLM to MCP servers - [Stagehand MCP](https://github.com/browserbase/mcp-server-browserbase) for browser access - [Firecrawl MCP](https://github.com/mendableai/firecrawl-mcp-server) for scraping - [Ragie MCP](https://github.com/ragieai/ragie-mcp-server) for multimodal RAG - [Graphiti MCP](https://github.com/getzep/graphiti/tree/main/mcp_server) as memory - [Terminal](https://github.com/wonderwhy-er/DesktopCommanderMCP) & [GitIngest](https://github.com/adhikasp/mcp-git-ingest) MCP ## Setup 1. **Install Dependencies**: ```bash uv sync ``` 2. **Environment Variables**: Create a `.env` file with your API keys: ```env OPENAI_API_KEY=your-openai-api-key FIRECRAWL_API_KEY=your-firecrawl-api-key RAGIE_API_KEY=your-ragie-api-key ``` 3. **Setup MCP Servers** Go to server.py and update the paths to the MCP servers according to your system. 3. **Run the App**: ```bash streamlit run mcp_streamlit_app.py ``` ## Usage 1. **Configure MCP Servers**: - Use the sidebar to enter your MCP server configuration in JSON format - Click "Load Example Config" to see a sample configuration - Click "Activate Configuration" to initialize the MCP client 2. **Chat with MCP Tools**: - Once configured, use the chat interface to interact with your MCP servers - Ask questions about available tools or request specific actions - The agent will use the appropriate MCP tools to respond ## Example Configuration ```json { "mcpServers": { "stagehand": { "command": "node", "args": ["/path/to/stagehand/dist/index.js"], "env": { "OPENAI_API_KEY": "your-api-key", "LOCAL_CDP_URL": "http://localhost:9222" } }, "firecrawl": { "command": "npx", "args": ["-y", "firecrawl-mcp"], "env": { "FIRECRAWL_API_KEY": "your-firecrawl-key" } } } } ``` ## 📬 Stay Updated with Our Newsletter! **Get a FREE Data Science eBook** 📖 with 150+ essential lessons in Data Science when you subscribe to our newsletter! Stay in the loop with the latest tutorials, insights, and exclusive resources. [Subscribe now!](https://join.dailydoseofds.com) [![Daily Dose of Data Science Newsletter](https://github.com/patchy631/ai-engineering/blob/main/resources/join_ddods.png)](https://join.dailydoseofds.com) --- ## Contribution Contributions are welcome! Please fork the repository and submit a pull request with your improvements.