Review triage

Skill Srltas/claude-skills/review-triage

My Claude Skills

Install

npx -y skills add Srltas/claude-skills --skill review-triage

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Triage the review comments on your pull request. Fetches a PR's reviews (inline code comments, review summaries, and PR-level comments, from humans and bots) and, for each, gives an easy Korean summary, judges whether it is valid by checking the actual code, then either drafts a fix plan (if valid) or a reasoned reply explaining why not (if not). Analysis and drafts only: it does not post replies or change code. Use when your PR has review feedback you want to understand and respond to. Triggers on phrases like 'PR 리뷰 정리해줘', '리뷰 코멘트 타당한지 봐줘', '리뷰 대응', 'triage my PR reviews', 'help me respond to PR review'.

SKILL.md

3.1 KB, 645 tokens by cl100k_base, as published. Nobody here has run it

Triage PR review comments

Pull the review feedback on your PR and, comment by comment, summarize it plainly, judge whether it holds up against the real code, and draft either a fix plan or a reasoned reply. Analysis and drafts only: it does not post anything or change code.

Step 1: Identify the PR

The current branch's PR by default, or a number/URL from $ARGUMENTS.

gh pr view --json number,url,headRefName,baseRefName   # current branch's PR

Step 2: Fetch the reviews

bash <skill-base-dir>/assets/fetch_reviews.sh [PR]

Merges inline code comments, review summaries, and PR-level comments (humans and bots, shown as (User) / (Bot)) with thread/reply info. Needs gh (authenticated) and jq.

Step 3: Judge each comment against the real code (grounding)

For every comment, open the referenced file:line and the PR diff and check whether the point is actually correct. Do not agree or dismiss from memory: read the code. Separate real issues from style opinions, misunderstandings, or already-handled cases. Use cubrid-manual / cmt-manual to confirm a fact when the comment hinges on one.

Step 4: Summarize and draft, per comment

For each comment, output (요약/판정/계획 or 답변 in Korean for the user):

요약: what the reviewer is asking, in plain Korean.
판정: 타당(O) / 부당(X) / 부분(△), with a one-line 근거 grounded in the code.
if 타당 → 수정 계획: which file and approach, briefly.
if 부당/부분 → 답변 초안: a reasoned reply to post. Write the reply in the reviewer's language (English for upstream repos such as Hibernate, Korean for CUBRID internal repos); keep the user-facing 요약/판정 in Korean.

If a reply needs to explain a non-trivial flow or relationship, an inline mermaid code block is an option (GitHub renders it in PR comments; nodes auto-size to text). Use it only when a diagram is clearer than a sentence.

Tone: open-source developer voice, plain and direct, not academic or formal-document. No em-dash (—): use commas, colons, parentheses, periods.

Step 5: Hand off (no posting, no fixing)

Give the user the per-comment triage. Do not post replies or edit code. If they want to reply, show the command only when they ask:

gh pr comment <N> --body "<reply>"    # PR-level reply
# inline thread reply:
# gh api repos/<owner>/<repo>/pulls/<N>/comments/<comment_id>/replies -f body="<reply>"

What ships with it: 1 file

4.1 KB alongside SKILL.md, 1 of them executable

assets/

fetch_reviews.shruns4.1 KB