本文へ移動
cccskills
無料GitHub で公開

reply-to-pr-conversation

Draft, confirm, and post a single conversational reply to GitHub PR conversation comments (issue comments). The reply addresses all tracked items in one natural-prose message. Use when the user asks to "reply to PR conversation", "post PR conversation replies", or "draft PR conversation messages".

インストール方法を見る

含まれるファイル(1)

  • SKILL.md3.1 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

Reply to PR Conversation

Draft a single reply that addresses a processed issue-comment list, confirm with the user, and post it as a new PR issue comment.

Step 1: Run /github-voice Skill

Run the /github-voice skill to load voice rules and the insider-vs-outsider detection.

Step 2: Compose the Reply

Use the processed-item list from conversation context. Each entry has: id, author, original comment body (to quote from selectively), category (fix, skip, answer, or clarify), and per-category payload.

Draft one reply that addresses every item as natural conversational prose. The output is a single piece of prose with flexible length and no rigid section structure.

Use the category to interpret each payload:

  • fix: payload is a commit SHA. Mention "fixed in <sha>" where it fits the flow, plus a brief note when the fix diverges from what the commenter suggested.
  • skip: payload is the skip reasoning. State it directly.
  • answer: payload is answer text prepared upstream. Integrate it as the implementer's own words.
  • clarify: payload is a user-directed question. Ask it as-is.

Quote selectively. Use > blockquotes only for the phrase being responded to. A single-topic reply quotes one sentence then responds; a multi-topic reply weaves quotes and responses together. Drop quotes entirely when @mentions plus context make the reply unambiguous.

Quote handling:

  • Strip leading > from quoted lines so nested blockquotes don't misattribute.
  • Replace fenced-code lines inside a quote with > [code snippet].
  • If the author login ends in [bot], omit the suffix from the @ mention.

Apply /github-voice rules. Match the conversation's length and tone. An acknowledgment with emoji is fine when the discussion calls for it; a multi-paragraph response is fine when the items warrant it.

Step 3: Confirm

Output the drafted reply as text for review:

**Draft comment**

<full comment body as it will be posted>

Then use AskUserQuestion to ask whether to post. Offer:

  • Post — post the comment
  • Cancel — skip posting

Step 4: Post the Comment

Auto-detect owner, repo, and PR number from the current branch if not provided. Write the drafted body to .turbo/pr/<pr_number>-comment.md with the Write tool, then post via the issue-comments REST endpoint:

gh api -X POST \
  "/repos/<owner>/<repo>/issues/<pr_number>/comments" \
  -F body=@.turbo/pr/<pr_number>-comment.md

Report the posted comment's URL.

Then use the TaskList tool and proceed to any remaining task.

Rules

  • Never close the PR, resolve a thread, or edit existing comments. Only post a new issue comment.
  • If the post fails, report the failure and leave the drafted body in the output so the user can post it manually.

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Apply findings by making the suggested code changes. Applies accepted verdicts, escalates ambiguous findings to the user, and offers to note genuine improvements for later. Use when the user asks to "apply findings", "apply fixes", "apply suggestions", "apply accepted findings", "fix the findings", or "apply the review results".

日本語の概要は準備中です。原文の説明を表示しています。

tobihagemann/turbo4092026年10月13日 更新

Audit the turbo skill lessons that /self-improve writes into this repo's .turbo/handoff/ from other projects, apply the ones the user approves, delete the consumed handoffs, commit and push, and update the installed Turbo skills. Use when the user asks to "read .turbo/handoff and let me know what you think", "read the handoff files and let me know what you think", "review the lesson handoffs", "apply the self-improve handoffs", or "process the turbo handoffs".

日本語の概要は準備中です。原文の説明を表示しています。

tobihagemann/turbo4092026年10月13日 更新

Assess project-wide structural technical debt: complexity hotspots, deprecated API usage, duplication clusters, architecture rot, and low-value tests. Ranks findings by impact and refactor effort into a report at .turbo/technical-debt.md. Use when the user asks to "assess technical debt", "find technical debt", "review technical debt", "what should we refactor", "find refactoring candidates", "where is the code rot", "what's our worst code", "find low-value tests", or "which tests can we delete". Analysis-only — does not modify code.

日本語の概要は準備中です。原文の説明を表示しています。

tobihagemann/turbo4092026年10月13日 更新

Shared changelog conventions and formatting rules referenced by /create-changelog and /update-changelog. Not typically invoked directly.

日本語の概要は準備中です。原文の説明を表示しています。

tobihagemann/turbo4092026年10月13日 更新

Shared changelog conventions and formatting rules referenced by $create-changelog and $update-changelog. Not typically invoked directly.

日本語の概要は準備中です。原文の説明を表示しています。

tobihagemann/turbo4092026年10月13日 更新

Run a non-interactive Claude Code print-mode call from Codex. Use when the user asks to "claude print", "ask claude", "run claude", "consult claude", or when a Codex Turbo skill needs Claude as an independent peer reviewer.

日本語の概要は準備中です。原文の説明を表示しています。

tobihagemann/turbo4092026年10月13日 更新

tobihagemann のスキルをすべて見る

このスキルの問題を報告する