draft-code-review-comment
MUST invoke when the user asks to draft, write, word, or turn an agreed code-review concern into a concise PR/code review comment in the current chat. Trigger phrases include "draft a review comment", "write a PR comment", "turn this into a code review comment", "comment for this change", "word this review feedback", and "make this a concise code review comment". Also use after a discussion about branch or PR changes when the user says the issue should be raised as review feedback. Do NOT invoke for performing a fresh review, checking whether an existing comment was addressed, or editing the code.
- License 许可
- Apache-2.0
- 最近更新
- 2026-11-14
- Stars
- ★ 1
- Forks
- 0
01 前言 Front Matter
- name
- draft-code-review-comment
- description
- MUST invoke when the user asks to draft, write, word, or turn an agreed code-review concern into a concise PR/code review comment in the current chat. Trigger phrases include "draft a review comment", "write a PR comment", "turn this into a code review comment", "comment for this change", "word this review feedback", and "make this a concise code review comment". Also use after a discussion about branch or PR changes when the user says the issue should be raised as review feedback. Do NOT invoke for performing a fresh review, checking whether an existing comment was addressed, or editing the code.
02 资源文件 1
- 参考文档agents/openai.yaml121 B
03 SKILL.md 正文
Draft Code Review Comment
Draft a paste-ready code review comment from the current conversation and the minimum additional code context needed to make the comment accurate.
Use this skill to convert an agreed concern into reviewer-facing wording. The output belongs in the chat, not in a file and not posted to GitHub unless the user separately asks for that.
Inputs
- Current conversation context, especially the agreed concern and desired change
- Optional branch, PR, diff, file path, line, test, or requirement context when needed for accuracy
- Optional tone constraints from the user, such as direct, softer, or very short
Workflow
- Identify the agreed review concern.
- Use the current conversation as the primary context.
- Confirm the target change, expected behavior, impact, and requested action.
- Do not reopen a broad review or look for unrelated findings.
- Gather only missing context.
- If the conversation is enough, draft immediately.
- If a factual detail is uncertain, inspect the relevant diff, file, tests, or branch context before stating it.
- Ask one concise clarification question only when the concern, target code, or desired change is still ambiguous after cheap inspection.
- Draft one concise comment.
- State the concrete problem, why it matters, and the requested change.
- Phrase the requested change as a direct instruction, not as a question.
- Keep it professional, direct, and high-confidence.
- Do not mention the AI discussion, internal reasoning, or that the comment was generated.
- Do not include severity labels, long background, or multiple alternatives unless the user asks.
- Avoid claims about missing tests, broken behavior, or runtime impact unless they were verified or clearly established in the conversation.
- Return the comment in chat.
- Provide the paste-ready comment as the main output.
- If there is an assumption or unresolved uncertainty, put it outside the comment block under
Assumption:orNeeds confirmation:. - Do not create files, modify code, or post the comment to an external system.
Output Format
Prefer this shape:
Suggested comment:
<one concise review comment>
If the user asks for alternatives, provide no more than three variants:
Direct:
<comment>
Softer:
<comment>
Short:
<comment>
Example
User context: "We agreed this retry helper logs the exception but still returns success. Draft the review comment."
Suggested comment:
This logs the retry failure but still lets the caller treat the operation as successful, so a failed request can continue through the workflow as if it completed. Please propagate the failure or return an explicit failed state here so the retry path remains observable.
Skill Chain
| Position | Details |
|---|---|
| Previous | Conversation, a prior review, or manual code discussion identifies an issue worth raising |
| This skill | Converts the agreed concern into one concise, paste-ready review comment in chat |
| Next | User posts the comment manually or requests a separate review |
Related Skills
pr-comment-addressed-check- determine whether an existing review comment has been addressed
04 同仓库更多 Skills
ui-ux-pro-max
UI/UX design intelligence for web and mobile. Local database with 79 searchable styles (88 catalog entries), 192 color palettes, 74 font pairings, 192 product types, 119 UX guidelines, 105 curated icon entries, 17 GSAP motion presets, and 25 chart types across 22 stacks (React, Next.js, Vue, Nuxt, Svelte, Astro, SwiftUI, React Native, Flutter, Tailwind, shadcn/ui, Jetpack Compose, Angular, Laravel, JavaFX, WPF, WinUI, Avalonia, Uno Platform, UWP, Three.js, and HTML/CSS). Use when designing, building, or reviewing UI: pages, components, color schemes, typography, layout, accessibility, animation, or data visualization.
figma-mcp
Implement or compare UI against Figma designs using an available Figma MCP integration. Use for Figma-backed implementation, design-token extraction, or visual-parity gap analysis; Figma MCP access is optional and must already be configured by the user.
improve-codebase-architecture
Find evidence-based, active-change architecture improvements using shared design and domain-modeling guidance. Use when a user wants to reduce architectural friction, deepen a module, or make a current change safer, more testable, and easier to navigate.
domain-modeling
Build and sharpen a project's domain model. Use when discussing codebase terminology, writing or editing GLOSSARY.md or CONTEXT.md, or recording or editing an ADR.
codebase-design
Provide the shared vocabulary and discipline for designing deep modules with clear interfaces and seams. Use when a user wants to design or improve a module, find deepening opportunities, choose a seam, or make code more testable and AI-navigable.
test-coverage-review
MUST invoke to review whether added or updated tests cover code changes in a diff. Checks new logic, branches, error paths, and regression risk, producing a change-to-test matrix. Use requirements-to-tests-traceability for acceptance-criteria mapping and orchestrated-review for bundled reviews.