When you're starting a README-first repo reproduction and need to know what's runnable before you touch anything, this is your scanner. It reads the README and project files, extracts documented commands, classifies them into inference, evaluation, and training buckets, then hands back a minimal reproduction plan. Think of it as the intake step before the real work starts. It won't set up environments, download assets, or run commands. It just maps the territory so an orchestrator or human can decide what to actually execute. The boundaries are strict by design: scan and plan only, no execution.
npx -y skills add lllllllama/rigorpilot-skills --skill repo-intake-and-plan --agent claude-codeInstalls into .claude/skills of the current project.
Select a file.
mattpocock/skills