verification-before-completion

Code completion verification requires running project-specific build, lint, and test commands, then inspecting their full output for warnings or partial failures before proceeding. This gated process ensures assertions like "done" or "fixed" are backed by evidence.

Evidence-first completion verification treats task completion as a gated action, requiring successful execution of build, test, and lint outputs before allowing a claim of done. It prevents premature completion claims by mandating actionable verification.

Build and test verification should be applied during commits, pull request creation, handoffs, and whenever you state done, fixed, or passing across software projects and configuration changes to ensure work is actually complete.

Pre-commit verification requires reading full command output to identify warnings, skipped tests, or partial failures, and mandates fixing these issues before proceeding to claim the work is complete.

Automated verification gating applies to configuration changes and bug fixes by selecting appropriate verification commands, executing them, and requiring all checks to pass before allowing completion claims or chaining to a self-review.

After successful build and lint verification, the workflow chains to a self-review process to continue the quality workflow, ensuring the verified code undergoes additional review before final completion.