verification-before-completion

Require fresh command output before reporting work as complete.

131|4|Updated Apr 1, 2026
One-click install
npx skills add https://github.com/AlphaLab-USTC/OhMyCode --skill verification-before-completion-alphalab-ustc
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: verification-before-completion
Source: https://github.com/AlphaLab-USTC/OhMyCode/tree/main/.agents/skills/superpowers/verification-before-completion
Command: npx skills add https://github.com/AlphaLab-USTC/OhMyCode --skill verification-before-completion-alphalab-ustc

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill prevents false completion claims by requiring fresh verification before you state that work is finished, fixed, passing, or ready to ship.

Core Features & Use Cases

  • Evidence-first completion: Forces you to run the proving command before asserting success.
  • Test and build verification: Ensures claims about passing tests, clean linting, successful builds, and bug fixes are backed by current output.
  • Agent output validation: Prevents blind trust in delegated work by requiring independent confirmation.
  • Use case: Before saying a regression is fixed or a PR is ready, verify the relevant command output and only then report the result.

Quick Start

Use the verification-before-completion skill when you are about to claim success, and first run the exact command that proves the claim.

Frequently Asked Questions about verification-before-completion

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

FAQPage Schema
How do I prevent false completion claims when reporting test or build results?

To prevent false completion claims, require fresh verification evidence by executing the proving command and reading its complete output before asserting that tests pass or builds succeed.

Why does my bug fix claim require fresh verification evidence before reporting success?

Bug fix claims require fresh verification evidence to ensure the regression is actually resolved by executing the proof command and matching the current output to the success assertion before reporting.

How to validate delegated agent output before stating work is finished?

To validate delegated agent output, execute the full proof command independently and read the complete result to confirm the delegated work matches the stated completion claim before reporting.

What's the best way to enforce evidence-first completion for linting and regression checks?

The best way to enforce evidence-first completion is to run the exact verification command for linting and regression checks, read the full output, and match the result to the claim before reporting success.

Does verification-before-completion work with delegated agent results and build verification?

Yes, it works with delegated agent results and build verification by requiring independent confirmation through executing the full proof command and matching the output to the claim before asserting success.