skills.rest
idehen-divine
idehen-divineCOMMUNITY
Explore all skills by this author →
Website𝕏@idehen_divineGitHub

check

Verify changes against specs and review diffs before merge.

Updated Sep 19, 2026
Search Tags:#code review#verification#acceptance criteria#quality gate#cross-model#git diff#pre-merge
One-click install
npx skills add https://github.com/idehen-divine/lumished --skill check-idehen-divine
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: check
Source: https://github.com/idehen-divine/lumished/tree/main/.ai/skills/check
Command: npx skills add https://github.com/idehen-divine/lumished --skill check-idehen-divine

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Green tests do not prove a feature actually works, and a model reviewing its own code shares its blind spots. This Skill closes both gaps before merge: it drives the real application to prove behavior against the spec, and runs a senior code review on a different model than wrote the code. ## Core Features & Use Cases - Runtime verification: The verify mode launches the real app, exercises the changed flow (UI, API, CLI, jobs), and reports pass or fail per behavior and per spec acceptance criterion, with cited evidence for every verdict. - Cross-model code review: The review mode spawns a reviewer subagent on a different model family than the author, ranks findings by severity (Blocker, Major, Minor, Nit), and writes them to docs/reviews/. - Spec conformance gate: When a governing spec exists, it confirms every acceptance criterion is met and every specced surface (page, route, table, migration) was actually built and applied. - Use Case: After finishing a feature branch, run verify to watch the new checkout flow work end to end against its spec, then run review before opening the PR so a fresh model catches correctness, security, and error handling issues in the diff. ## Quick Start Ask the AI to run the check skill in verify mode to prove the current change works against its spec, or in review mode for a fresh-model code review of the diff.

Frequently Asked Questions about check

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

FAQPage Schema
How do I verify a feature actually works before merging?▼

Run the check skill in verify mode. It scopes the change from git, launches the app using the project's own run command, exercises the changed flow, and reports pass or fail per behavior with cited evidence such as screenshots, request responses, or command output.

How to get a code review from a different AI model?▼

Run the check skill in review mode. It detects which model wrote the code, spawns a reviewer subagent on a contrasting model family, and writes severity-ranked findings to docs/reviews/. You can force a reviewer with the 'with <model>' argument.

Does the check skill modify my code?▼

No, both modes are read only on code. Verify runs the app and reports failures pointing to /debug or /develop, while review only writes a findings file. Neither mode edits the code under examination.

What happens if the app cannot be started during verification?▼

The verify mode reports affected behaviors as blocked rather than passing them. It never emits a pass without recorded evidence, so a missing browser tool, credentials, or a failing build results in an honest blocked verdict naming what is needed.

Can I use the check skill on projects without a test suite?▼

Yes. The review mode recognizes a none-by-design test signal where the project gates on typecheck plus runtime verification, and it will not raise missing-test findings. Verify mode serves as the primary safety net for such projects.

Related Skills

Explore More Software Engineering Skills→
XinheLIUXinheLIU

spec

Create a git worktree and feature branch with a draft spec.md from a template.

Community
Intermediate
Base-AnaliticaBase-Analitica

test-driven-development

Enforces red-green-refactor test-driven development workflow before writing production code.

Official
Intermediate
erfanzarerfanzar
27

requesting-code-review

Automate pre-commit code review with static checks, linting, and auto-fixes.

Community
Advanced
Open-PawsOpen-Paws
3

code-review

Automate layered code review for pull requests in CI pipelines.

Official
Advanced