CodeLens — Demo
CodeLens is an offline, AI-powered semantic code search engine that helps developers understand and navigate their codebase using natural language. Instead of relying on keyword matching, CodeLens analyzes the meaning of code by converting it into vector embeddings, allowing you to search with queries like “where is authentication handled?” and instantly find the most relevant parts of the code.
Built with a local-first architecture using Ollama and a lightweight vector database, CodeLens runs entirely on your machine — no internet, no API keys, and full privacy. Integrated directly into VS Code, it provides a seamless experience where search results are instantly clickable, taking you straight to the exact file and line of code.
Think of it as Ctrl+F, but intelligent — designed to make large codebases easier to explore, understand, and work with.