What problem does it solve? Acceptance gates in design docs often certify broken or vacuous behavior: tests that pass for the wrong reason, negative assertions that can never fail, or gates referencing test files and seams that do not exist. This Skill audits every gate before implementation so the design doc cannot green-light broken features. ## Core Features & Use Cases - Honesty checks: Attacks each gate for vacuous negatives, timing-window vacuity, poll-interval vs timeout conflicts, unprovable real-time halves, routing/dispatch false negatives, port races, and process-global env poisoning. - Feasibility verification: Greps the actual test project to confirm named test files, patterns (FakeTimeProvider settles, host-shape pins, E2E factories), and production seams exist, and verifies factual claims like "no test churn" by counting literal constructions. - Structured report: Emits a verdict summary, findings table with MUST-FIX/SHOULD-FIX/NIT severities, per-criterion HONEST/FEASIBLE verdicts, corrected TDD order per work package, and missed test cases. - Use Case: Before implementing a BackgroundService idle watchdog design with 13 acceptance gates, run the audit to catch that a "shutdown within 5s" gate is unsatisfiable with a 1-minute poll loop and that a negative pin passes whether or not routing exists. ## Quick Start Audit the acceptance gates in my design doc for honesty and feasibility before I start implementing the work packages.
