MCP Server POC
by TarcisioPhilips
A proof-of-concept implementation of a Model Context Protocol (MCP) server for enhancing AI assistant capabilities with custom tools and resources. It provides custom tools and resources to AI assistants like Claude and others that support the MCP standard.
Last updated: N/A
What is MCP Server POC?
This project is a Model Context Protocol (MCP) server that provides custom tools and resources to AI assistants. It includes tools for documentation search, web crawling, and integration with Google Search API, enabling AI assistants to access and utilize external information.
How to use MCP Server POC?
To use the MCP server, first clone the repository, set up a virtual environment with Python 3.11, and install the required packages from requirements.txt. Then, configure the environment variables in a .env file. Run the server using python main.py. Finally, integrate it with Cursor IDE by configuring the ~/.cursor/mcp.json file with the correct path to main.py and restarting Cursor.
Key features of MCP Server POC
Documentation search for LangChain, LlamaIndex, and OpenAI
Web crawling capability
Integration with Google Search API
Windows binary mode fix for stdio transport
Integration with Cursor IDE
Use cases of MCP Server POC
Enhancing AI assistant knowledge with up-to-date documentation
Providing AI assistants with real-time information via web crawling
Enabling AI assistants to perform web searches
Integrating custom tools into AI assistant workflows
Developing custom tools for AI assistants
FAQ from MCP Server POC
What is the Model Context Protocol (MCP)?
What is the Model Context Protocol (MCP)?
The Model Context Protocol (MCP) is a standard that allows AI assistants to access and utilize external tools and resources.
What libraries are supported for documentation search?
What libraries are supported for documentation search?
The server supports documentation search for LangChain, LlamaIndex, and OpenAI.
What API key is required?
What API key is required?
A Serper API key is required for the Google Search API integration. You must set the SERPER_API_KEY environment variable.
How do I integrate the server with Cursor IDE?
How do I integrate the server with Cursor IDE?
Configure the ~/.cursor/mcp.json file with the correct path to main.py and restart Cursor completely.
What do I do if I get 'Failed to create client' or 'Client closed' errors on Windows?
What do I do if I get 'Failed to create client' or 'Client closed' errors on Windows?
Make sure to use the binary mode fix included in the server, use the absolute path with double backslashes in the mcp.json configuration, try running the MCP server directly to see if it produces any error output, and completely exit Cursor (including terminating any background processes via Task Manager) before restarting.