Skill Pantry · Gaiya365

Quality

Verification Before Completion

Back completion claims with the command output that actually proves them.

QualityMITChecked Oct 18, 2026

Our review

This short skill is a final evidence gate. It asks an agent to name the claim, run the relevant check, read its output, and report exactly what the result supports. It distinguishes a successful command from an assumption based on old logs, another agent’s report, or a test run before the last edit.

Use it before saying a fix is complete, tests pass, or a build works. Skip a formal checklist when there is no factual completion claim to make, though a quick inspection is still sensible. The method is deliberately strict and may add time to simple work. That is useful only if the selected check really covers the change; passing an unrelated test gives false confidence.

The folder has only SKILL.md and no included executable code or network call. The verification commands it prompts an agent to run come from the host project, not from this skill. Inspect those commands and their side effects before running them, and keep the final report tied to fresh output.

For installation, open the linked maintainer folder and read its current SKILL.md, license, and supporting files. Copy the complete folder to a skills location your agent documents, or use an installer your agent supports. Begin with a small, low-risk task and confirm that the agent loaded the intended version. Our install guide covers those checks.

Source and license

Maintainer: Jesse Vincent / obra. Checked: October 18, 2026. The maintainer controls this repository; the current files may differ from the version we reviewed.

Repository ↗ · SKILL.md ↗ · MIT license ↗

← Browse all skills