From ai-review
Presents git changes via ai-review desktop app for optional human review and feedback after task completion. Handles uncommitted diffs, gg stacks, feature branches, commits; applies parsed feedback.
How this skill is triggered — by the user, by Claude, or both
Slash command
/ai-review:show-changesThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
You have completed a task and want to show the human what you did. Open ai-review so they can look over the changes and optionally leave feedback. This is less formal than a full code review — the framing is "here's what I did" rather than "please review my code."
You have completed a task and want to show the human what you did. Open ai-review so they can look over the changes and optionally leave feedback. This is less formal than a full code review — the framing is "here's what I did" rather than "please review my code."
Determine what to show. Check these in priority order to pick the right air invocation:
a. Uncommitted changes exist (staged or unstaged, check via git status --porcelain):
Run air --wait --json with no branch/commit flags. This shows the working directory diff.
b. All changes committed, on a gg stack branch — the current branch matches the user/name pattern (no -- in the name part) and .git/gg/config.json exists:
Read the gg base branch from config (parse defaults.base, fall back to main). Run:
air --wait --json --commits <gg-base>..HEAD
This shows the full stack diff, matching what ai-review displays when opening a gg stack.
c. All changes committed, on a feature branch (not main/master):
Determine the base branch (main or master, whichever exists). Run:
air --wait --json --branch <base-branch>
d. All changes committed, on main/master itself:
Identify the first commit you made during this session. If you can determine the exact commit, use that hash directly. Otherwise, fall back to the first unpushed commit: run git log origin/main..HEAD --oneline --reverse (or origin/master) and take the first commit hash. Run:
air --wait --json --commit <first-relevant-commit-hash>
e. None of the above: Run air --wait --json with no flags (empty diff — the app handles it).
The air --wait --json command opens the ai-review desktop app and blocks until the human submits. They may add comments or submit with no comments. The output is structured JSON.
Handle the response.
If the human submitted comments: Parse the JSON feedback. The output has the shape:
{
"format": "ai-review.feedback/v1",
"context": { "mode": "...", ... },
"comments": [
{ "id": "...", "file": "...", "startLine": 10, "endLine": 12, "side": "old|new", "text": "...", "createdAt": "..." }
]
}
Address each comment: read the referenced file and lines, make the requested change. If a comment is unclear, use your best judgment.
If the human submitted with no comments: (empty comments array) Acknowledge that they've seen the changes and move on.
Summarize. If changes were made, provide a concise summary listing each comment and what you did to address it.
air is not found, tell the user to install via brew install --cask mrmans0n/tap/ai-review (macOS) or build from source at https://github.com/mrmans0n/ai-review.npx claudepluginhub mrmans0n/ai-review --plugin ai-reviewProvides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.