isc-playbook-mcp-server

# ISC Playbook MCP Server Model Context Protocol (MCP) server for searching and browsing the IBM ISC Playbook with hybrid search capabilities. ## Features - **Hybrid Search**: Combines semantic (dense vectors) and keyword (sparse vectors) search using Milvus Lite - **1,358 Indexed Pages**: Complete IBM ISC Playbook content (1,091 documents + 266 folders) - **Natural Language Queries**: Search using conversational questions - **GitHub Copilot Integration**: Access playbook directly from VS Code ## Architecture - **Vector Database**: Milvus Lite (embedded, no external service needed) - **Embeddings**: sentence-transformers (all-MiniLM-L6-v2, 384-dim) - **Search Algorithm**: RRF (Reciprocal Rank Fusion) for result merging - **MCP Framework**: FastMCP (Python SDK) ## Installation ### Option 1: Using uvx (Recommended) The easiest way to use this MCP server is with `uvx`: ```bash # Install and run directly with uvx uvx isc-playbook-mcp-server ``` ### Option 2: From PyPI ```bash # Install from PyPI pip install isc-playbook-mcp-server # Run the server isc-playbook-mcp-server ``` ### Option 3: From Source (Development) 1. **Clone the repository**: ```bash git clone git@github.ibm.com:kirtijha/isc-playbook-mcp-server.git cd isc-playbook-mcp-server ``` 2. **Install with uv**: ```bash uv pip install -e . ``` Or with pip: ```bash python -m venv .venv source .venv/bin/activate # On macOS/Linux pip install -e . ``` ## Usage ### MCP Server Configuration Add to your MCP settings file (e.g., `~/.config/mcp/mcp.json` or VS Code's MCP settings): #### Using uvx (Recommended): ```json { "mcpServers": { "isc-playbook": { "command": "uvx", "args": ["isc-playbook-mcp-server"] } } } ``` #### Using installed package: ```json { "mcpServers": { "isc-playbook": { "command": "isc-playbook-mcp-server" } } } } ``` ### Available Tools 1. **search_playbook**: Natural language search across a…