Skip to content
Get packs

Verify Changes Before Claiming Done

Makes the agent prove a change works with tests or direct checks before saying it is complete

Source Download skill
Repository
Yeachan-Heo/oh-my-claudecode
Stars
40,512
Licence
MIT
Files
1 file

What it does

When installed, this skill changes how the agent wraps up coding work. Instead of saying a feature, fix, or refactor is finished because it should work, the agent first identifies the exact behavior to prove, runs existing tests, then typechecks or builds, then runs narrow direct commands, and falls back to described manual steps if needed. It then reports only what was actually verified, including failures and anything left unverified.

Great for

  • Developers who want evidence that AI-written code changes actually work before merging
  • Teams that rely on existing test suites and want the agent to run them consistently
  • Anyone who has been burned by an agent declaring success without running anything

Skip it if

  • Quick exploratory prototyping where verification is not needed yet, since the skill adds a required evidence step
  • Projects with no tests, build tooling, or runnable commands, because the skill can only report that verification is not possible and adds little value

Try asking

  • “Verify that my fix for the login redirect bug actually works before you tell me it's done”
  • “Check this refactor of the payment module with the existing tests and a typecheck, and tell me what passed and what didn't”
  • “Confirm the new date-parsing function works and show me a concise summary of the commands you ran”
Step-by-step guide: how to use Verify Changes Before Claiming DoneInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 1 original files, the MIT licence and an attribution note.

    verify.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip verify.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip verify.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Give the agent a clear description of the behavior you expect so it knows exactly what to prove
  • Ask for a concise evidence summary if you want to avoid long raw logs
  • Check the agent's report for the unverified section, since some changes can only be checked manually

What's inside

  • SKILL.md

Redistributed unmodified from Yeachan-Heo/oh-my-claudecode under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.