Self-hosted alternatives to analyze-project

A

analyze-project

Rigor Analyze / Rigor Audit read-only skill

analyze-project is a read-only agent skill from the rigorpilot-skills collection for understanding a deep learning research repository before anyone changes it. It is the Rigor Analyze and Rigor Audit skill, and its installed name is kept for compatibility. The skill guides the assistant to read a repository and build a map of its model structure, its training entry points, its inference entry points and how its configuration files relate to each other. It can also make conservative suggestions about where a change would likely be inserted and point out implementation patterns that look suspicious, but it only suggests those and does not change code. The skill leaves the model room for project-specific reasoning, so it is a guide to the analysis rather than a fixed script. It is meant for a researcher who has inherited a repository, wants to know where training starts, or needs to see how the pieces fit before running anything. It is explicitly for read-only analysis and not for heavy execution, so it is a safe first step before experiments. It should not be used when the user wants a run or a fix. The skill passes Gen Agent Trust Hub, Socket and Snyk on skills.sh, and it has about 319,000 installs. Install it with npx skills add https://github.com/lllllllama/rigorpilot-skills --skill analyze-project.