verification-before-completion

Verification before completion requires running actual test commands, reading output, checking exit codes, and confirming zero failures. Only then can you claim success with corroborating evidence instead of assumptions.

Running fresh verification prevents false success reports that lead to broken trust, shipped bugs, and wasted rework. Evidence-based validation replaces confidence with concrete proof from actual test runs and build outputs.

Before completion, execute the full verification command—tests, builds, and linting—read all outputs, validate exit codes show success, count failures as zero, and provide output evidence. This enforces quality gates and prevents premature claims.

No. Verification before completion eliminates premature claims by requiring actual command execution and proof. Confidence without evidence leads to bugs shipping; this Skill forces concrete validation before any success declaration.

By requiring fresh verification evidence—test output, build logs, linting results—before claiming completion, this Skill acts as a quality gate. It blocks commits and PRs without proof, catching failures before they ship.