Content
# AI Project Explorer
AI Project Explorer is a Python-based Model Context Protocol (MCP) server that enables AI assistants to discover and interact with GitHub repositories through standardized tools.
Instead of directly calling the GitHub API, AI clients such as Claude Desktop, Cursor, or ChatGPT can invoke MCP tools to explore repositories, retrieve project information, and answer questions about software projects.
This project demonstrates how MCP enables AI applications to interact with external systems through reusable, discoverable tools.
## Why I Built This
I wanted to understand the Model Context Protocol (MCP) by building a real tool
The goal was to learn:
- How MCP clients discover tools
- How MCP servers expose capabilities
- How AI assistants invoke external functions
- How to integrate third-party APIs through MCP
- The difference between local (STDIO) and remote (HTTP) MCP transports
AI Client
(Claude / Cursor / ChatGPT)
│
▼
Python MCP Client
│
▼
AI Project Explorer MCP Server
│
▼
GitHub REST API
│
▼
Public GitHub Repositories
## Run The Inspector
Start the MCP Inspector with the latest Inspector package:
```bash
npm run inspector
```
This runs:
```bash
npx @modelcontextprotocol/inspector@latest .venv/bin/python server.py
```