05-review — Claude Code skill: what it does, how to install
05-review inspects git diffs or implementation plans to confirm changes are ready to ship, pairing every issue with a fix.
MIT ★ 571 updated 2026-10-29
/plugin marketplace add ai-driven-dev/framework
/plugin install aidd-dev@aidd-frameworkWhat 05-review does
The skill evaluates pending changes across three axes: code hygiene (review-code), alignment with plan criteria (review-functional), and overall necessity of edits (review-relevancy). Once all checks run, a final step returns one of three verdicts: approve, changes-requested, or blocked.
Evaluation is strictly static: the assistant never executes the target application. This tool points out flaws and names solutions, but it leaves codebase editing to the user and will not audit an entire repository.
What it helps with
- determine whether a git diff is safe to ship
- match changed lines against requirements from the plan
- spot unnecessary files or code leaked into a change
- get actionable corrections for every reported issue
Example prompts for the agent
- Run 05-review on the current git diff
- Check our implementation plan using 05-review
- Review the staged changes across all review axes
How to install 05-review
1Plugin marketplace
/plugin marketplace add ai-driven-dev/framework/plugin install aidd-dev@aidd-frameworkThis installs the whole aidd-dev plugin with all its skills
2Skills CLI (Vercel)
npx skills add ai-driven-dev/framework --skill 05-review -a claude-code3Manually: put the files in place
~/.claude/skills/05-review/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/05-review/
1Skills CLI (Vercel)
npx skills add ai-driven-dev/framework --skill 05-review -a codex2Manually: put the files in place
.agents/skills/05-review/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add ai-driven-dev/framework --skill 05-review -a cursor2Manually: put the files in place
.cursor/skills/05-review/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install ai-driven-dev/framework plugins/aidd-dev/skills/05-review2Manually: put the files in place
.github/skills/05-review/SKILL.md1Skills CLI (Vercel)
npx skills add ai-driven-dev/framework --skill 05-review -a gemini-cli2Platform CLI
gemini skills install https://github.com/ai-driven-dev/framework --path plugins/aidd-dev/skills/05-reviewManually: put the files in place
.devin/skills/05-review/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add ai-driven-dev/framework --skill 05-review -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about 05-review
How to install 05-review in Claude Code?
Add the catalog with /plugin marketplace add ai-driven-dev/framework, then execute /plugin install aidd-dev@aidd-framework.
Does 05-review run the project or test suite?
No. The procedure is strictly static, which means running the application is explicitly prohibited.
Will 05-review fix the detected problems?
It identifies each defect and pairs it with a concrete remedy, but it does not modify the source files.
Source text
SKILL.md · MIT
--- name: 05-review description: Reviews a change or a diff and says whether it is ready to ship, naming every problem with its fix. Use when the user wants a change checked before it ships. Not for fixing what the review finds, and not for auditing a whole codebase. argument-hint: diff | plan --- # Review ```mermaid flowchart LR all([all axes]) --> prepare one([one axis named]) --> prepare prepare --> review-code & review-functional & review-relevancy --> finalize finalize --> approve([approve]) finalize --> changes([changes-requested]) finalize --> blocked([blocked]) changes -.->|a later run| all blocked -.->|a later run| all ``` ## Actions Run the flow above, reading only the next action file. | Action | Does | | --- | --- | | prepare | resolve the diff and open the round | | review-code | grade the changed lines on clean code | | review-functional | trace the diff against the plan's criteria | | review-relevancy | judge whether the change belongs | | finalize | judge the round and check its shape | ## Transversal rules - Ask which entry the caller wants when the request does not say. - Review statically: never run the app.
© AI-Driven Dev · MIT · ai-driven-dev/framework