File Rank MCP Server

# File Rank MCP (Model Context Protocol) Server A TypeScript-based tool for ranking files in your codebase by importance and providing summaries with dependency tracking. ## Overview This tool helps you keep track of which files in your project are most important, giving them a rating between 0 (least important) to 10 (most important). It also tracks file dependencies and provides summaries, helping you understand the codebase structure. ## Features - Rank files on a scale from 0-10 (10 being most important) - Track and visualize file dependencies (which files import which) - Real-time file system monitoring for automatic updates - Get summaries of files (with placeholder for LLM integration) - Query rankings for individual files, directories, or the entire codebase - Self-documenting API with discovery endpoints - Simple stdio JSON interface ## Setup 1. Clone this repository 2. Make the build script executable: ```bash chmod +x build.sh ``` 3. Run the build script: ```bash ./build.sh ``` This will: - Check for Node.js installation - Install project dependencies - Compile TypeScript code - Generate MCP configuration - Create a detailed log file in `/tmp/file-rank-mcp_*.log` 4. Move the generated `mcp.json` to your project's `.cursor/` directory to use with Cursor AI ## Usage The tool communicates via JSON over stdio using the Model Context Protocol (MCP). It automatically monitors your project directory for changes and updates the file tree accordingly. ### Running the Server ```bash node dist/mcp-server.js ``` ### Supported Languages The server automatically detects dependencies in the following languages: - Python: `import` and `from ... import` statements - C/C++: `#include` directives - Rust: `use` and `mod` statements - Lua: `require` statements - JavaScript/TypeScript: `import` statements ## How It Works The tool maintains a real-time tree structure of your codebase, stored in `file-tree.json`. This file is automatically updated whenever changes are detected in your project directory. ### Data Structure The file tree is stored in a JSON structure: ```json { "path": "/path/to/directory", "name": "directory_name", "isDirectory": true, "children": [ { "path": "/path/to/file.ts", "name": "file.ts", "isDirectory": false, "dependencies": ["/path/to/dependency.ts"], "importance": 8 } ] } ``` ### Dependency Detection The tool analyzes source code to detect dependencies: - Scans file contents for language-specific import patterns - Resolves relative imports to absolute paths - Verifies that imported files exist - Calculates importance based on dependency count ### File Importance Calculation Files are assigned importance scores (0-10) based on: - Number of dependencies (files that import this file) - Position in the dependency tree - File type and location ## Technical Stack - TypeScript/Node.js - Model Context Protocol SDK (v1.7.0) - Chokidar for file system monitoring - ES2020 target with CommonJS modules ## Development To modify or extend the project: 1. Make changes to the TypeScript files in the `src` directory 2. Run the build script to compile and update configuration: ```bash ./build.sh ``` 3. Run the server: ```bash node dist/mcp-server.js ``` The build script will: - Install any new dependencies - Compile TypeScript code - Generate updated MCP configuration - Create a detailed log file for troubleshooting ## Integration with LLMs The current implementation includes a placeholder for LLM integration. In a production environment, the server could be extended to: - Generate file summaries using LLMs - Analyze code complexity and maintainability - Provide intelligent suggestions for file organization - Generate documentation based on code structure ## License This project is licensed under the GNU General Public License v3 (GPL-3.0). See the [LICENSE](LICENSE) file for the full license text.