← ClaudeAtlas

spec-review-gauntletlisted

Use before implementation to review a Spec Bundle, design direction, or execution plan for readiness, traceability, scope discipline, multi-lane design risks, and TDD/verification feasibility. DO NOT USE FOR: implemented-code review or small already-frozen edits.
funky-eyes/best-copilot · ★ 10 · AI & Automation · score 82
Install: claude install-skill funky-eyes/best-copilot
# Spec Review Gauntlet Use this skill as the pre-implementation gate for MEDIUM/LARGE work and for high-risk customization workflow changes. It prevents a plan from entering implementation before requirements, design, task boundaries, and review ownership are ready. ## Inputs Read only the minimum necessary evidence: - user-provided paths, current editor file, attachments, and priority files - current `requirements.md`, `design.md`, `tasks.md`, or equivalent approved plan - relevant owner files and already-read context supplied by PM - changed files, search hints, and reference files when supplied Do not expand to broad repository search unless a specific missing evidence category is identified. ## Readiness Checks Check these dimensions before implementation: - `completeness`: main path, edge cases, rollback/cleanup, and closeout are covered. - `correctness`: requirements, design, tasks, defaults, and non-goals agree. - `placeholder_scan`: no `TBD`, `TODO`, template residue, or unresolved choice remains. - `scope_discipline`: file boundaries, non-goals, and forbidden approaches are explicit. - `existing_code_leverage`: the plan reuses existing modules, helpers, tests, and ownership boundaries. - `traceability`: each requirement maps to implementation tasks, tests/checks, and review lanes. - `tdd_testability`: new behavior or bug fixes can be proven with a failing test or minimal reproducible check. - `task_granularity`: tasks are small enough to execute and review in