Can this skill handle regression testing?

Yes, it specifically includes patterns for Red-Green testing, ensuring a test fails when the fix is removed and passes when it is restored.

What is the primary purpose of this Claude Code skill?

It prevents the AI from making premature or 'hallucinated' success claims by requiring it to run real terminal commands and verify the output before reporting a task as finished.

When does the 'Iron Law' of verification apply?

The rule applies before any variation of success claims, expressions of satisfaction, or moving to commit/push code to a repository.

Does it trust third-party agent reports?

No, the skill explicitly mandates independent verification of VCS diffs and changes even when delegated agents report success.

How does this improve the development workflow?

It reduces rework and 'broken' code submissions by ensuring that tests, linters, and builds have actually passed in the current environment before the AI suggests moving forward.

Verification Before Completion

Name: Verification Before Completion
Author: AlabamaMike

byAlabamaMike

0•

安全与测试

Enforces a strict evidence-based protocol requiring empirical verification before any task is marked as complete or successful.

This skill establishes an 'Iron Law' for AI-assisted development, preventing premature claims of success by requiring fresh, empirical evidence from verification commands. It acts as a rigorous gatekeeper that mandates running tests, builds, and linters—and confirming their full outputs—before any commit, pull request, or task closure. By eliminating vague assertions like 'should work' and replacing them with verified data, it ensures high-quality code output and maintains developer trust through absolute transparency.

主要功能

01Comprehensive validation checklists for requirements, builds, and lints

02Evidence-based reporting protocol that identifies, runs, and reads command outputs

03Red-Green-Refactor patterns for verified regression testing

040 GitHub stars

05Mandatory execution of verification commands before any success claims

06Strict prevention of subjective language like 'should' or 'probably'

使用场景

01Preventing broken builds by mandating exit-code verification before code commits

02Validating complex feature requirements against a line-by-line checklist before PR creation

03Ensuring bug fixes are truly resolved through rigorous red-green test cycles

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add alabamamike/copilot-agents verification-before-completion

For use in Claude.ai and ChatGPT

Download Skill