cqs

What it does

cqs is a code intelligence and retrieval-augmented generation (RAG) tool designed for AI agents. It provides semantic code search, call graph analysis, impact assessment, type dependency tracking, and intelligent context assembly through single tool calls. The platform supports 54 programming languages and 29 different chunk types, enabling comprehensive code understanding and navigation.

How to set up

cqs operates as a local-first daemon service with GPU acceleration support. Installation typically involves cloning the repository from GitHub and running the daemon locally. Once started, the service exposes query endpoints that can be called by AI agents or integrated into development workflows. Configuration allows customization of language support and chunk type preferences.

Key features

• Semantic code search with 73% R@5 and 85% R@20 retrieval accuracy
• Call graph generation and analysis for code navigation
• Impact assessment and type dependency tracking
• Multi-language support across 54 languages
• Ultra-fast daemon queries (3-19ms response times)
• GPU-accelerated processing for enhanced performance
• Local-first architecture ensuring privacy and control
• RAG-optimized context assembly for AI agents