Build a non-trivial change from an approved plan.md. Use after /spec to capture a baseline, execute dependency-aware subtasks, record evidence and deviations, run independent /vet and /exercise gates, measure the delta, and stop for human ship approval without committing or publishing automatically. Don't use for…
Surface hidden requirements and architectural unknowns before planning non-trivial work. Use when a request is ambiguous, spans unfamiliar code, has meaningful tradeoffs, or needs a grounded blindspot pass before /spec. Produces an unknowns-map.md artifact and stops without implementing. Don't use when you already…
Verify a completed change through real user behavior after automated tests. Use as /craft's independent behavioral gate for web, desktop, CLI, installer, documentation, or static-site workflows. Drives concrete scenarios, captures evidence for every assertion, and returns PASS or FAIL. Don't use for static code review…
Stage and commit code changes with Conventional Commits format. Use when: code changes are complete and ready to commit, after review approval, user says 'save', 'commit', 'ship it'. Don't use when: tests are failing or work is still in progress.
Render a Claude Code plan (PlanViewer), a task result, or a recap of a branch/commit/PR diff (ResultViewer) as a single-file, self-contained HTML artifact with anti-fatigue UX. Use when the user asks for an HTML artifact, a plan/result/recap view, or to "show this as HTML". Not for plain markdown/text or one-liners.
Scan project and generate a customized CLAUDE.md. Detects stack, test commands, and conventions. Use when onboarding to a new project or user says 'init', 'setup claude', 'generate CLAUDE.md'.
Scan a codebase for accumulated slop, technical debt, dead code, and refactor opportunities. Default is recent changes only. Use when user says 'scan for slop', 'find tech debt', 'what needs cleanup', or before a refactor sprint. Don't use as the pre-PR gate — that is /vet (lens D).
Plan a non-trivial change before implementation. Use when work needs repository research, locked interfaces, security invariants, dependency-aware subtasks, executable acceptance criteria, and explicit approval. Consumes unknowns-map.md when available, writes plan.md, and stops without coding. Don't use for trivial…
Independently review a working tree, branch, pull request, or implemented plan before ship. Use for a material pre-merge gate that checks acceptance criteria, regressions, security boundaries, project consistency, cleanup quality, and active counterexamples in fresh context. Returns SHIP or HOLD with evidence. Don't…