Analyze codebase

Build an inventory of what's in the repo so other skills don't have to re-discover it. The output is a structured summary, not a written file — the orchestrator passes it as context to the next sub-skill.

What to scan, in order

Stop at the first finding for each section unless the section says otherwise. Don't recurse into node_modules/, .git/, dist/, build/, .turbo/, .next/, __pycache__/, .venv/, target/.

1. Top-level fingerprint

ls -la at repo root + git log --oneline -10. Capture:

• Repo name (from package.json:name, pyproject.toml:[project].name, Cargo.toml:[package].name, or git remote).
• Default branch (git symbolic-ref refs/remotes/origin/HEAD or fall back to main/master).
• Code host (parse git remote get-url origin — github.com / gitlab / bitbucket / self-hosted).
• Existing AGENTS.md / CLAUDE.md / .cursorrules / .github/copilot-instructions.md / CONTRIBUTING.md — read these in full. Existing conventions are gospel.

2. Languages + package managers