DeepContext is an MCP (Model Context Protocol) server designed to enhance the capabilities of AI coding agents like Claude Code and Codex CLI. It addresses the limitations of traditional text-based search (e.g., `grep`), which often floods agents with irrelevant information, especially in large code repositories. By implementing symbol-aware semantic search, DeepContext intelligently identifies and returns only the most relevant code chunks, leveraging AST-based parsing and a hybrid search approach combining vector similarity and BM25 with Jina reranking. This precise context provision leads to reduced token usage, faster search results, fewer AI hallucinations, and ultimately, more efficient and accurate coding assistance.
主要功能
01Symbol-aware semantic search for precise code understanding
02Reduced token usage by returning only relevant code chunks
03Fast search through pre-indexed codebase data
04AST-based parsing for deep code structure analysis
05Hybrid search with reranking for optimal relevance
0610 GitHub stars
