Show HN: Vesalius, an on-device code graph for Claude Code (github.com)

Vesalius, a new on-device code graph tool optimized for Claude Code, was recently introduced, enhancing the functionality of AI coding assistants. Unlike traditional methods that require multiple search and open commands to source code information, Vesalius creates a comprehensive map of your repository that maintains relationships between files, functions, and features. This innovation allows Claude Code to retrieve answers about code efficiently, significantly cutting down the time and tokens used during queries, as evidenced by a benchmark that reduced token consumption by nearly 50% while maintaining answer quality. This development is significant for the AI/ML community as it streamlines the interaction between AI assistants and codebases, enabling faster and more accurate responses to complex coding questions. With features like dependency checks, feature mapping, and health diagnostics, Vesalius not only assists AI agents but also provides developers with a robust tool for exploring their projects. Importantly, it preserves user privacy by keeping all operations local without uploading code to servers, which could be vital for developers working on sensitive projects. This tool represents a substantial leap towards smarter and more efficient coding practices, leveraging the power of on-device AI to enhance productivity.