verification-before-completion

Verification requires running your full test command (e.g., npm test, pytest), pasting the complete output, and confirming the results pass before stating completion. This command-first approach provides evidence that backs every completion claim and prevents false positives from incomplete test runs.

Assuming success without running verification masks hidden failures in linting, builds, or test suites. Fresh verification evidence—captured output from the actual command—confirms your changes are genuinely sound and prevents shipping broken code into production or review.

Paste the full, untruncated output from your verification command—tests, CI checks, or linting—so the complete result is visible and auditable. Partial output can hide failures; full output proves the command ran to completion and all checks passed.

State the failure clearly with the evidence from your verification command output. Document what failed, fix the root cause, re-run the full command, and only after fresh verification passes should you claim the issue resolved or the PR ready.

Verification applies across all development workflows: unit tests, integration tests, linting, builds, and CI checks. Any task where you claim completion must be backed by running the designated verification command and pasting evidence that confirms the result.

Yes. Running verification commands before submitting a PR catches test failures, lint errors, and build issues early. Enforcing verification as a gate prevents incomplete work from reaching review and reduces rejection cycles.