Srag addresses the common limitation of AI coding assistants by enabling them to understand and leverage your specific codebase, patterns, and conventions. By indexing all your repositories, srag allows AI assistants to perform semantic searches, providing instant access to relevant implementations, patterns, and architectural context. This integration not only helps in reusing existing code and maintaining consistent style but also enhances debugging, onboarding, and cross-project knowledge transfer, ultimately leading to more accurate code suggestions and reduced token usage through its Model Context Protocol (MCP) server.
Key Features
01Multi-project indexing with cross-project search and pattern analysis
02Tree-sitter based AST-aware code chunking for 9 programming languages
03Model Context Protocol (MCP) server for AI tool integration
04Semantic and full-text hybrid code search with reranking
05Prompt injection detection and secret redaction
0615 GitHub stars
