verification-before-completion

Verify test results and build output before claiming task completion.

Updated May 28, 2026
One-click install
npx skills add https://github.com/kyo0221/claude-config --skill verification-before-completion-kyo0221
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: verification-before-completion
Source: https://github.com/kyo0221/claude-config/tree/main/skills/verification-before-completion
Command: npx skills add https://github.com/kyo0221/claude-config --skill verification-before-completion-kyo0221

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill eliminates the risk of making unverified completion claims (such as stating tests pass, bugs are fixed, or work is done) without running actual verification, which leads to broken code, wasted rework, and broken trust with collaborators and stakeholders.

Core Features & Use Cases

  • 5-Step Verification Gate: Enforces a mandatory process of identifying the correct verification command, running it fresh, reading full output, confirming the claim matches evidence, and only then making a success claim.
  • Common Failure Reference: Lists exactly what verification is required for common claims (e.g., test passes require 0 failures in fresh test output, not previous runs) and what evidence is insufficient.
  • Guardrail Tools: Includes red flag warnings for common signs of unverified claims and rationalization prevention to counter common excuses for skipping verification steps.
  • Use Case: Before committing code or opening a PR, use this Skill to run your full test suite, confirm 0 failures from the fresh output, and only then state the work is complete, avoiding shipping broken code to production.

Quick Start

Use the verification-before-completion skill to confirm all tests pass with fresh output before claiming your code changes are ready to commit.

Frequently Asked Questions about verification-before-completion

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I verify code changes before claiming a task is complete?

To verify code changes before claiming completion, identify the correct verification command, run it fresh, review the full output and exit codes, and only state completion if directly supported by the evidence. This prevents shipping broken code.

Why do my pull requests contain unverified completion claims?

Unverified completion claims occur when developers state tests pass or bugs are fixed without running fresh verification commands. This leads to broken code and wasted rework; enforcing a 5-step verification gate prevents this issue.

What is the best way to prevent shipping broken code to production?

The best way to prevent shipping broken code is to enforce workflow guardrails requiring fresh full test suite execution and output review before committing code or opening a pull request, ensuring zero failures are confirmed.

How do I validate bug fix confirmation before committing code?

To validate bug fix confirmation before committing, run the specific verification command fresh, read the complete output and exit codes, and ensure the results directly match your bug fix claim without relying on previous test runs.

Does verification before completion work for delegating tasks to agents?

Yes, verification before completion applies to delegating tasks to agents. It requires agents to execute fresh full verification commands, review complete output, and only express satisfaction with work output supported by verification evidence.

What are common red flags of unverified completion claims?

Common red flags of unverified completion claims include relying on previous test runs instead of fresh output, skipping verification steps through rationalization, and stating work is done without checking complete command exit codes.