Skills Prompts

Popular categories

Agent skill · Addy Osmani's Agent Skills

/code-review-and-quality Five-axis code review with clear severity labels

This skill has the agent review any change, whether written by you, another model or itself, for correctness, readability, architecture, security and performance. It reads the tests first, checks them by flipping a condition, labels each comment as required, critical, optional, nit or FYI, proposes a concrete restructuring for structural problems and reviews dependency upgrades one package at a time.

Works with

  • Claude Code
  • Any agent via the skills CLI
  • Codex
  • Gemini CLI
  • Antigravity CLI
  • Command Code

Used forCode review

When to use it

  • A pull request or local diff needs a verdict before merge.
  • Another agent or model wrote code you have not examined.
  • A bulk dependency bump or a refactor claims to make things cleaner.

How it works

  1. 1
    Understand intent

    The agent works out what the change is meant to do and which spec or task it implements before reading code.

  2. 2
    Test the tests

    It reads the tests first, then inverts a condition the change adds and reruns the suite; a test that stays green is reported as a gap.

  3. 3
    Walk five axes

    Each changed file is checked for correctness, readability, architecture, security and performance, including size of the diff and of the file.

  4. 4
    Label and order findings

    It drops findings that callers or tests rule out, separates old defects from new regressions, tags severity and puts the highest-impact issues first.

  5. 5
    Check verification

    It confirms which tests and builds were run and lists any now-unused code, asking before deleting it.

What it steers the agent away from

  • Approving with LGTM and no evidence of review
  • Refactors that move complexity around without removing any
  • Merging a mass version bump without reading changelogs or the lockfile diff

Install this skill

It installs with the Addy Osmani's Agent Skills pack.

Claude CodeAdds the skills plus nine lifecycle slash commands; use /agent-skills:review where a built-in command shares the name.
/plugin marketplace add addyosmani/agent-skills
/plugin install agent-skills@addy-agent-skills
Any agent via the skills CLIInstalls all skills into Claude Code, Cursor, Codex, Copilot, Cline and many other agents.
npx skills add addyosmani/agent-skills
CodexNeeds Codex CLI v0.122 or later; skills are then called with @ in chat.
codex plugin marketplace add addyosmani/agent-skills
codex plugin add agent-skills@agent-skills
Gemini CLIInstalls the skills for auto-discovery.
gemini skills install https://github.com/addyosmani/agent-skills.git --path skills
Antigravity CLIInstalls as a native plugin; some releases need the namespaced skills invoked directly.
agy plugin install https://github.com/addyosmani/agent-skills.git
Command CodeThe first form installs per project, the second for all projects.
cmd skills add addyosmani/agent-skills
cmd skills add addyosmani/agent-skills --global

Commands come from the author’s README; check it for changes before you run them.

Questions

Is a review with no findings acceptable?

Yes. The skill says zero findings is a valid result and the agent should not invent issues to fill the report.

How large should a change be?

It treats about 100 changed lines as ideal, about 300 as acceptable for one logical change, and around 1000 as a signal to split.

More guides from Addy Osmani's Agent Skills

Also in this pack

All 24 skills →

Summary by trendhub. The skill itself lives in addyosmani/agent-skills/skills/code-review-and-quality by Addy Osmani (MIT); read the full, current text there.