speckit-checklist_skill

Overview

This skill generates a custom requirements-quality checklist for a feature based on the user's description and any available spec/plan/tasks. It frames checklist items as "unit tests for requirements" that validate clarity, completeness, consistency, measurability, and coverage. The output is a new, uniquely named checklist file placed under the feature's checklists directory.

How this skill works

The skill inspects the user's prompt and optional feature artifacts (spec.md, plan.md, tasks.md) and extracts domain signals and risk indicators. It asks up to three targeted clarifying questions when the request is ambiguous, loads only the relevant sections from the feature directory, then produces a grouped checklist whose items evaluate the written requirements (not implementation). Each item is numbered CHK001+ and includes quality dimension tags and spec references or [Gap]/[Ambiguity] markers where appropriate.

When to use it

• When you need a requirements-quality checklist tailored to a specific feature or domain
• Before a PR review to validate that requirements are ready for implementation
• When drafting or refining specs to surface missing or ambiguous requirements
• When reviewers or stakeholders require measurable acceptance criteria
• When preparing release gates that must assess requirement completeness and traceability

Best practices

• Provide brief focus and scope in the initial request to reduce clarification questions
• Attach or point to spec/plan/tasks so the skill can reference sections and add traceability markers
• Accept up to three clarifying questions to ensure checklist relevance and precision
• Prefer identifying high-risk areas (security, rollback, accessibility) up front so items prioritize them
• Use the generated checklist as a living artifact — produce a new file per run and consolidate later if needed

Example use cases

• Generate a UX requirements checklist to confirm visual hierarchy, interaction-state definitions, and accessibility coverage
• Create an API-spec checklist to surface missing error modes, acceptance criteria IDs, and dependency assumptions
• Produce a release-gate checklist that emphasizes rollback, migrations, and measurable non-functional thresholds
• Draft a security-focused checklist that checks for threat-model assumptions, auth boundaries, and traceability
• Create a reviewer-facing checklist for PRs that maps items to spec sections and flags ambiguities