Why does it flag words like 'should' or 'probably'?

Those words indicate a lack of evidence. This skill requires certainty through execution, replacing assumptions with verified command outputs.

What is the 'Iron Law' of this skill?

The Iron Law states that no completion claims can be made without providing fresh verification evidence from a command run in the current context.

Does this skill trust reports from other AI agents?

No. It requires independent verification of VCS diffs and actual state changes rather than relying on an agent's success message.

How does this skill handle regression testing?

It follows a strict red-green pattern: verify the test fails with the bug, apply the fix, and then verify the test passes to prove the fix works.

When should I use the Verification Before Completion skill?

It should be applied before any variation of success claims, committing code, creating PRs, or moving to the next task in your workflow.

Verification Before Completion

Name: Verification Before Completion
Author: tianyuanztm

bytianyuanztm

0•

Security & Testing

Enforces a strict evidence-first policy for all task completion claims to ensure code reliability and build trust.

This skill implements a rigorous 'Iron Law' for software development, preventing premature completion claims by requiring fresh, documented verification evidence before any work is declared finished. It serves as a critical gate function that mandates identifying specific verification commands, executing them in full, and analyzing the actual output before committing code or creating pull requests. By eliminating 'should-work' assumptions and hollow assertions, this skill ensures that tests pass, builds succeed, and requirements are met with documented proof, significantly reducing regressions and developer rework.

Key Features

01Anti-rationalization checks to prevent 'should work' or 'confident' excuses

02Structured TDD Red-Green cycle enforcement for regression testing

030 GitHub stars

04Independent validation of changes delegated to other AI agents

05Evidence-based verification via fresh command execution and output analysis

06Mandatory gate function for all status and completion claims

Use Cases

01Ensuring rigorous bug fix validation through verified regression testing

02Preventing broken code or failing builds from being pushed to a repository

03Confirming all project requirements and linter rules are satisfied before PR creation

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add tianyuanztm/myplan verification-before-completion

For use in Claude.ai and ChatGPT

Key Features

01Anti-rationalization checks to prevent 'should work' or 'confident' excuses

02Structured TDD Red-Green cycle enforcement for regression testing

030 GitHub stars

04Independent validation of changes delegated to other AI agents

05Evidence-based verification via fresh command execution and output analysis

06Mandatory gate function for all status and completion claims

Use Cases

01Ensuring rigorous bug fix validation through verified regression testing

02Preventing broken code or failing builds from being pushed to a repository

03Confirming all project requirements and linter rules are satisfied before PR creation

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add tianyuanztm/myplan verification-before-completion

For use in Claude.ai and ChatGPT