.agents/skills/gh-pr-review/SKILL.md
Automated code review for local branches, PRs, commits, and files. Detects review mode from arguments and routes to the appropriate review flow — either quick single-agent review with interactive fix selection, or multi-agent deep review with risk-based auto-fix.
Cherry Studio-specific review rules live in
references/cherry-review-guidance.md. Target review flows must load that file
for code, mixed, architecture-doc, and project-skill reviews so reviewers can
apply DataApi, service-boundary, renderer hook, React, UI, and type-contract
checks without relying on memory. That reference also defines which internal
docs, internal skills, external skills, and official websites to consult for
each changed area; load only the relevant subset.
All user-facing text matches the user's language. Use the runtime's interactive dialog tool for questions and option selection when one is available; otherwise ask one concise plain-text question and wait for the reply. Do not invent a tool or syntax the runtime does not expose. For interactive multi-select: ≤4 items → one question. >4 items → group by priority or category (each group ≤4 options), then present all groups in one prompt.
Run pre-checks, then match the first applicable rule top-to-bottom:
git branch --show-current → record whether on main/master.git status --porcelain → record whether uncommitted changes exist.| # | Condition | Action |
|---|---|---|
| 1 | $ARGUMENTS is diag | → references/diagnosis.md |
| 2 | $ARGUMENTS is a PR number or URL containing /pull/ | → references/pr-review.md |
| 3 | Agent teams NOT supported | → references/local-review.md |
| 4 | Uncommitted changes exist | → references/local-review.md |
| 5 | On main/master branch | → references/local-review.md |
| 6 | Everything else | → Question below |
Each → means: Read the target file and follow it as the sole remaining
instruction. Ignore all sections below. Do NOT review from memory or habit —
each target file defines specific constraints on how to obtain diffs, apply
fixes, and submit results.
Priority rule: user intent (Rule 1, 2, 6) takes priority over working-tree state (Rule 3, 4, 5). A PR URL or PR number always goes to
references/pr-review.mdeven when the working tree is dirty or the current branch ismain/master— those state conditions only apply when the user did not specify a review target.
Ask a single question: "Agent Teams is available (multiple agents working in parallel). Enable multi-agent review with reviewer–verifier adversarial mechanism and auto-fix?" Provide 4 options:
| Option | Description |
|---|---|
| Teams + auto-fix low & medium risk (recommended) | Multi-agent review; auto-fix most issues, only confirm high-risk ones (e.g., API changes, architecture). |
| Teams + auto-fix low risk | Multi-agent review; auto-fix only the safest issues (e.g., null checks, typos, naming). Confirm everything else. |
| Teams + auto-fix all | Multi-agent review; auto-fix everything. Only issues affecting test baselines are deferred. |
| Single-agent + manual fix | Single-agent review; interactively choose which issues to fix afterward. |
| Option | → | FIX_MODE |
|---|---|---|
| Teams + auto-fix low & medium risk (recommended) | references/teams-review.md | low_medium |
| Teams + auto-fix low risk | references/teams-review.md | low |
| Teams + auto-fix all | references/teams-review.md | full |
| Single-agent + manual fix | references/local-review.md | — |
Pass $ARGUMENTS to the target file. For teams-review, also pass FIX_MODE
(low / low_medium / full).