Local Search MCP Server

# Usage Guide ## Installation ### Quick Start (Recommended) ```bash # Install globally from npm npm install -g local-search-mcp # Or run directly with npx (no installation needed) npx -y local-search-mcp ``` ### Development Setup ```bash # Clone and build from source git clone https://github.com/PatrickRuddiman/local-search-mcp.git cd local-search-mcp npm install npm run build ``` ## Configuration ### For NPM Package Installation Add to Claude Desktop's `claude_desktop_config.json`: ```json { "mcpServers": { "local-search": { "command": "npx", "args": ["-y", "local-search-mcp"], "env": { "MCP_DATA_FOLDER": "/optional/custom/data/path", "MCP_DOCS_FOLDER": "/optional/custom/docs/path" } } } } ``` ### For Source Installation Add to Claude Desktop's `claude_desktop_config.json`: ```json { "mcpServers": { "local-search": { "command": "node", "args": ["/absolute/path/to/local-search-mcp/build/index.js"], "env": { "MCP_DATA_FOLDER": "/optional/custom/data/path", "MCP_DOCS_FOLDER": "/optional/custom/docs/path" } } } } ``` ## Basic Usage ### 1. Index Documentation ```typescript // Index a GitHub repository await fetch_repo({ repoUrl: "https://github.com/microsoft/vscode-docs" }); // Or download individual files await fetch_file({ url: "https://raw.githubusercontent.com/expressjs/express/main/README.md", filename: "express-docs.md" }); ``` ### 2. Search Content ```typescript await search_documents({ query: "authentication API", options: { limit: 5, minScore: 0.7 } }); ``` ### 3. Get Detailed Content ```typescript await get_file_details({ filePath: "/docs/auth.md", chunkIndex: 2, contextSize: 3 }); ``` ## Advanced Usage ### Job Monitoring ```typescript // Check background processing status await get_job_status({ jobId: "job_abc123" }); // List all active jobs await list_active_jobs(); ``` ### File Management ```typescript // Remove indexed files await remove_file({ filePath: "/docs/old-file.md" }); ``` ## Claude Desktop Integration The server provides semantic search capabilities to Claude Desktop. Ask questions naturally: ``` "How does JWT authentication work in this project?" "What are the API error response patterns?" "Where are the database migration files?" ``` Claude will automatically search your indexed documentation and provide contextual answers. ## Best Practices - **Chunk Size**: Default 1000 chars works well for most documentation - **File Types**: Automatic detection supports 10+ formats (MD, TXT, JSON, etc.) - **Concurrent Processing**: Background jobs prevent blocking MCP responses - **Memory Usage**: Embedding models require 2GB+ RAM See [API Reference](api/) for complete specifications and [Performance Guide](performance/) for optimization tips.