verification-before-completion

Use when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always

Safety Notice

This listing is imported from skills.sh public index metadata. Review upstream SKILL.md and repository scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "verification-before-completion" with this command: npx skills add futuregerald/futuregerald-claude-plugin/futuregerald-futuregerald-claude-plugin-verification-before-completion

Verification Before Completion

Never declare work complete without running verification commands and confirming their output. Evidence before assertions, always.

Core Rule

Before claiming ANY of the following, you MUST run the relevant verification commands and confirm their output:

  • "Tests pass"
  • "The fix works"
  • "Build succeeds"
  • "Everything is green"
  • "Ready to commit/push/merge"
  • "Done" / "Complete" / "Finished"

What Counts as Verification

Acceptable evidence:

  • Running the test suite and seeing the output (pass count, zero failures)
  • Running the build command and seeing "build succeeded" or equivalent
  • Running typecheck/lint and seeing zero errors
  • Reproducing the specific scenario that was broken and confirming it now works
  • Running the specific test that was failing and confirming it now passes

NOT acceptable evidence:

  • "I think this works based on the code I wrote"
  • "This should fix it" (without running anything)
  • "The logic looks correct" (code reading alone)
  • Committing without running tests
  • Declaring done because the implementation matches the plan

Workflow

  1. Before claiming tests pass: Run the test command. Read the output. Confirm zero failures.
  2. Before claiming a fix works: Reproduce the original failure scenario. Confirm it no longer fails.
  3. Before claiming the build succeeds: Run the build command. Read the output. Confirm no errors.
  4. Before committing: Run tests + typecheck + lint. All must pass.
  5. Before creating a PR: Run the full verification suite. Confirm CI-equivalent checks pass locally.

Red Flags

Stop and verify if you catch yourself:

  • Saying "I believe this works" without having run it
  • Committing immediately after writing code
  • Skipping tests because "it's a small change"
  • Declaring a bug fixed without reproducing the original issue
  • Saying "tests should pass" instead of "tests pass — here's the output"
  • Trusting that your code is correct because it looks right

The Mantra

"Show me the output." Every claim of completion must be backed by command output you have actually seen in this session.

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

Automation

dispatching-parallel-agents

No summary provided by upstream source.

Repository SourceNeeds Review
Automation

agent-browser

No summary provided by upstream source.

Repository SourceNeeds Review
Automation

verification-before-completion

No summary provided by upstream source.

Repository SourceNeeds Review