Make your agent check the real thing before it says done

Before your agent says it is done, it runs the real program from outside and reads what it prints, instead of trusting a compile, a cached screenshot or its own word.

Free, to use and to change. What it can see: only the chat you add it to.

Its short name: prove-it. Source on GitHub.

Install

Pick the one you use.

Claude Code

/plugin marketplace add mbsi-labs/skills
/plugin install prove-it@skills

Changelog

2026-10-16 (6f94f8b): Start mbsi-skills with five short-name skills adapted from pstack

Licence: MIT.

Read the skill itself

Unverified work has unknown correctness. Check the real thing directly.

Run the real program from outside and read what it prints (a demonstration, definitions.md). Not mtimes, not a cached screenshot, not an agent's word.
Check liveness directly (the process, the listener), not state derived from it.
When a check fails, suspect the way you looked before you suspect the system.
Script the check when you can, so a reviewer reruns it instead of trusting you. Put the command and its output in the hand-back.
A script that proves a fix is a demonstration, not a new test. Add a test only for a promise no step checks yet (code-standards.md).
Visual results go to the user, shown inline. Their eyes are the check.