How does this skill handle regression testing?

It follows a strict red-green-red cycle: you must write the test, verify that it fails without the fix, apply the fix, and then verify that the test passes.

What is the primary goal of the Verification Before Completion skill?

The primary goal is to ensure that no work is claimed as complete, fixed, or passing without fresh, documented evidence from actual command outputs.

Why is this skill necessary for AI agents?

AI agents can sometimes hallucinate success or overlook side effects; this skill forces a 'trust but verify' workflow that ensures the code actually works in the real environment.

What does the skill consider a 'Red Flag'?

Red flags include using words like 'should', 'probably', or 'seems to', and expressing satisfaction ('Great!', 'Done!') before running verification commands.

Verification Before Completion

Name: Verification Before Completion
Author: franroa

byfranroa

0•

Security & Testing

Enforces a strict evidence-based workflow that requires running verification commands before claiming any task is finished.

This skill implements a rigorous 'Iron Law' for AI-driven development: evidence before assertions. It prevents premature claims of success by mandating that the AI execute fresh verification commands—such as tests, builds, and linters—and confirm the output before reporting a status or creating a pull request. By eliminating vague language like 'should work' and requiring a red-green verification cycle, it significantly reduces regressions and ensures that every code change is backed by hard data, building a foundation of trust and reliability in the development process.

Key Features

01Systematic requirement validation through line-by-line checklists

020 GitHub stars

03Detection and prevention of non-committal language like 'should' or 'probably'

04Independent verification of changes when delegating to sub-agents

05Strict red-green cycle enforcement for regression testing and bug fixes

06Mandatory execution of verification commands before any success claim

Use Cases

01Ensuring a project builds and passes all linting rules before committing code

02Confirming that all requirements in a complex task list have been met with evidence

03Validating that a bug is actually fixed by running the specific failing test case

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add franroa/chezmoi verification-before-completion

For use in Claude.ai and ChatGPT

Download Skill