Trusty Search provides hybrid semantic, lexical, and relational code search across named indexes served by a single local daemon. It exposes tools to locate code definitions, explore conceptual flows, find technical debt like TODOs, and run synthesized Q&A over the codebase with citations. Reach for it when you need precise symbol lookup or deep codebase exploration without manual file grepping.
Key Features
Hybrid retrieval using BM25, vector similarity, and knowledge graphs
Query type routing for definition, usage, conceptual, and bug debt searches
LLM-powered Q&A over code indexes with citations
Index management and single-file reindexing capabilities
Privacy & Security
Data Collection
This tool follows industry-standard security practices and only collects data necessary for functionality.