Provides AI assistants with local documentation search capabilities from Git repositories and local directories.
Sponsored
Local Docs empowers AI assistants by providing a robust system for indexing and searching documentation from various sources. It efficiently pulls content from GitHub repositories using sparse checkouts and local directories, storing it in a SQLite database with FTS5 for fast, full-text search. This allows AI to dynamically discover, index, and retrieve relevant information, ensuring answers are within context window limits through intelligent token budgeting.
主要功能
011 GitHub stars
02Sparse checkout for efficient Git repository indexing
03FTS5 full-text search with BM25 ranking
04Token budgeting to fit results within AI context windows
05Asynchronous background indexing and automatic refresh
06Markdown chunking to preserve document structure
使用案例
01Enabling AI assistants to discover and index relevant documentation dynamically.
02Searching external GitHub repositories or local directories for specific information.
03Providing AI agents with up-to-date, context-aware documentation for answering queries.
