Gortex is a powerful code intelligence engine designed to revolutionize how AI coding agents interact with codebases. It meticulously indexes entire repositories, constructing a rich in-memory knowledge graph that maps out files, symbols, imports, call chains, and type relationships across 33 programming languages. By providing highly optimized and contextually relevant information through a single `smart_context` call, Gortex drastically cuts down token usage by up to 94% compared to traditional file reads, making AI coding agents like Claude Code, Kiro, and Copilot far more efficient and effective. It exposes this intelligence via a CLI, an MCP Server, and a web UI, offering a comprehensive suite of tools for semantic search, multi-repo management, and agent feedback loops.
Key Features
01Indexes code into a queryable in-memory knowledge graph
02Supports multi-repo workspaces with cross-repo symbol resolution
03Analyzes 33 programming languages including Go, Python, JavaScript, and C++
04Offers hybrid BM25 + vector semantic search with offline embedding options
05Tracks token savings per call and session for AI agent efficiency
062 GitHub stars
Use Cases
01Performing semantic code analysis and graph traversal for developer insights and impact assessment
02Optimizing AI coding agents (e.g., Claude Code, Copilot) for efficient context retrieval and reduced token usage
03Managing and querying multi-repository projects with unified cross-repo intelligence
