Docsearch
Provides local-first, hybrid semantic and keyword search for various document types and Confluence pages through the Model Context Protocol.
About
Docsearch is a powerful local-first document search and indexing system designed for hybrid semantic and keyword search. It allows users to quickly find information across a multitude of sources, including local files (code, documentation, and PDFs) and Confluence spaces. Built with the Model Context Protocol (MCP), it seamlessly integrates with AI assistants like Claude Code/Desktop, enabling them to access your codebase, research materials, and other documentation efficiently and securely. It emphasizes data privacy by keeping API keys and sensitive data on your machine.
Key Features
- Hybrid Search: Combines full-text search with vector similarity for optimal results across indexed content.
- Multi-Source Indexing: Indexes local files (code, docs, PDFs) and content from Confluence spaces.
- AI-Powered Image Search: Generates descriptions for images like diagrams and screenshots, making them searchable.
- Real-time Updates: Includes file watching for automatic re-indexing of local file changes.
- MCP Integration: Seamlessly integrates with AI assistants like Claude Code and other MCP-compatible tools.
- 3 GitHub stars
Use Cases
- Empowering AI assistants to access personal documentation, codebases, and research materials securely.
- Searching local code repositories, project documentation, and PDF research papers efficiently.
- Centralizing search capabilities across multiple Confluence spaces and local files for rapid information retrieval.
Sponsored