verification-before-completion

Require fresh command output and exit status before asserting completion.

Updated Apr 24, 2026
One-click install
npx skills add https://github.com/Cain-Ish/claude-code-plugin --skill verification-before-completion-cain-ish
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: verification-before-completion
Source: https://github.com/Cain-Ish/claude-code-plugin/tree/main/skills/verification-before-completion
Command: npx skills add https://github.com/Cain-Ish/claude-code-plugin --skill verification-before-completion-cain-ish

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It prevents false completion claims by forcing verification evidence (fresh command output and exit status) before you say work is done, fixed, passing, or ready to ship.

Core Features & Use Cases

  • Evidence-first completion gate: Requires identifying the specific command that proves the claim, running the full verification command, reading the full output, and only then asserting status.
  • Anti-rationalization guardrails: Rejects “should/probably/seems”, satisfaction phrasing, reliance on partial checks, and trusting agent success reports without independent verification.
  • Common verification mappings: Aligns typical claims (tests, linter, build, bug fixes, regression, completion) with the exact kind of evidence that must be produced.

Quick Start

Before you say “done”, “fixed”, “passes”, or “ready to merge,” run the full test/build/lint command that proves it, confirm the output and exit code, and only then report the result with evidence.

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 in CI and pull requests?

To prevent false completion claims in CI and pull requests, require fresh verification evidence like command output and exit status before asserting any work is done, passing, or ready to merge.

What is evidence-first verification for release readiness?

Evidence-first verification for release readiness is a process requiring the execution of full test, build, and lint commands to produce current output and correct exit codes before reporting a passing or fixed status.

How do I validate agent-assisted work without trusting success reports?

To validate agent-assisted work without trusting success reports, run independent verification commands to inspect the full output and confirm the exit status before claiming the task is complete or fixed.

How to stop rationalization during code reviews and testing?

To stop rationalization during code reviews and testing, reject phrases like 'should' or 'seems' and require explicit verification mappings that align completion claims with exact test, linter, or build evidence.

Does verification before completion work without external dependencies?

Yes, verification before completion works without external dependencies by acting as a reasoning guardrail that forces you to run existing testing, linting, and CI commands to prove readiness before claiming completion.