What problem does it solve? Teams and individuals accumulate scattered documentation that is hard to search meaningfully; keyword search misses context and cloud-based search raises privacy concerns. This Skill runs a local-first MCP server that ingests documents, chunks them intelligently, and makes them searchable by meaning. ## Core Features & Use Cases - Hybrid Semantic Search: Combines full-text and vector search across all documents using an embedded Orama database with parent-child chunking for richer context. - Document Management: Add, list, retrieve, and delete documents, or drop .txt, .md, and .pdf files into an uploads folder for batch processing. - Optional Gemini AI Search: Ask natural-language questions across selected documents with conversation history when a GEMINI_API_KEY is configured. - Use Case: A developer drops API guides and runbooks into the uploads folder, runs process_uploads, then queries "rate limiting configuration" from their MCP client and retrieves the exact section plus surrounding context. ## Quick Start Add the documentation server to my MCP client configuration using npx @andrea9293/mcp-documentation-server and index my markdown files for semantic search.
