Agent Skills › doodledood/manifest-dev › review-pr-holistic

review-pr-holistic

GitHub

该技能用于代码审查,整合审查者发现与PR历史,执行去重、前提归纳和范围限定,按调用方语气重写评论,生成最终发布内容。

claude-plugins/manifest-dev-tools/skills/review-pr-holistic/SKILL.md doodledood/manifest-dev

Trigger Scenarios

需要合并多个审查维度的缺陷与判断结果 需对已有评论进行去重、过滤和语气统一

Install

npx skills add doodledood/manifest-dev --skill review-pr-holistic -g -y
More Options

Non-standard path

npx skills add https://github.com/doodledood/manifest-dev/tree/main/claude-plugins/manifest-dev-tools/skills/review-pr-holistic -g -y

Use without installing

npx skills use doodledood/manifest-dev@review-pr-holistic

指定 Agent (Claude Code)

npx skills add doodledood/manifest-dev --skill review-pr-holistic -a claude-code -g -y

安装 repo 全部 skill

npx skills add doodledood/manifest-dev --all -g -y

预览 repo 内 skill

npx skills add doodledood/manifest-dev --list

SKILL.md

Frontmatter
{
    "name": "review-pr-holistic",
    "description": "Called by review-pr after its reviewer fleet and judgment pass return: consolidates their findings against the PR's history into the set of comments worth posting — pruning what the PR already covers, merging duplicates, subordinating defects under premise questions, bounding to the reviewed range, and rewriting every body in the caller's voice. Not for direct use.",
    "user-invocable": false
}

review-pr-holistic

Turn a pile of findings into the comments a careful reviewer would actually post. This skill edits and consolidates; it generates nothing new and posts nothing itself.

Receives. Fleet findings — severity-graded defect findings from the reviewer dimensions, with dimension and severity retained, Low findings filtered by the caller except those from defect-class — and judgment findings — a distinct non-severity class, each { trigger, concrete evidence, author-facing question }, already synthesized to one question per root; the PR history (all comments and threads on the PR, including ours from any prior pass, the author's recent commit messages on the branch, the PR description); bundle context for each linked PR (diff, description, top-level conversation — no inline review comments from linked PRs); the manifest if present; the reviewed range for this invocation — the code the fleet's defect findings are bounded to; the caller's voice profile — its rules for how a posted comment reads; and any truncation the caller did, to carry forward.

Does, in this order:

  • Prunes duplicate concerns already covered by existing threads, findings refuted by evidence, and defect findings whose entire objection is to an explicit intentional requirement. A stated intention establishes scope; evidence establishes behavior. Retain a premise question unless the PR already answers that specific concern. An author's claim alone does not refute a demonstrated failure.
  • Dedupes across reviewers: merge near-duplicates into one comment when they raise the same underlying concern.
  • Premise-subsumption. When a judgment finding questions the necessity or existence of a surface, subordinate the fleet's findings on that same surface under the premise question — drop the low-value ones, and fold any that survive beneath it ("if you keep X: also a, b") rather than listing them as independent peers. Retain a defect-class finding, including Low, when its remediation obligation remains unmet; a non-blocking premise question cannot discharge it. Never hard-delete a severe independent defect: the judgment finding is non-blocking, so a defect that matters even if the surface stays is still surfaced on its own.
  • Bounds surfaced defect findings to issues introduced or exposed by the reviewed range, unless the range is the full PR diff — which covers a first review and a repeat review of an unchanged head alike. This keys on the range, never on whether we have reviewed before: bounding a repeat pass to the delta since our last review would have it read the whole PR and then surface nothing. A repair in the reviewed range exposes its enabling boundary and affected sites even when their lines are unchanged; an unrelated later delta does not reopen the old repair. Judgment findings are exempt: they are whole-PR by construction and stay exempt from the drop-Low step too — they surface whenever their evidence fired and the PR does not already cover the point.
  • Anchors each surviving finding to exactly one of inline file:line (default), file-level (a whole-file concern), or PR-level (cross-cutting, no specific anchor).
  • Rewrites every comment body and any drafted reply in the supplied voice profile.
  • Omits a summary header by default — adds one only when there is a real overall take the per-comment list misses (one short sentence, voice-compliant, no boilerplate).

Returns. The comments to post (anchor plus voice-compliant body), the summary header text if any, truncation notes, and a brief dropped-findings tally with the dominant reasons.

Version History

  • d76241f Current 2026-09-28 09:21

    优化低严重度缺陷处理逻辑,增强基于审查范围的边界控制,确保修复暴露的依赖关系被正确保留。

  • 3fc997a 2026-09-22 22:25

Same Skill Collection

.claude/skills/auto-optimize-prompt/SKILL.md
.claude/skills/compress-prompt/SKILL.md
.claude/skills/harden-task-file/SKILL.md
.claude/skills/learn-from-session/SKILL.md
.claude/skills/optimize-prompt-token-efficiency/SKILL.md
.claude/skills/sync-claude-code-plugins/SKILL.md
.claude/skills/sync-tools/SKILL.md
claude-plugins/manifest-dev-tools/skills/babysit-pr/SKILL.md
claude-plugins/manifest-dev-tools/skills/eli5/SKILL.md
claude-plugins/manifest-dev-tools/skills/handoff/SKILL.md
claude-plugins/manifest-dev-tools/skills/prompt-engineering/SKILL.md
claude-plugins/manifest-dev-tools/skills/re-pitch/SKILL.md
claude-plugins/manifest-dev-tools/skills/review-pr-judgment/SKILL.md
claude-plugins/manifest-dev-tools/skills/review-pr-thread-verify/SKILL.md
claude-plugins/manifest-dev-tools/skills/review-pr/SKILL.md
claude-plugins/manifest-dev-tools/skills/review-prompt/SKILL.md
claude-plugins/manifest-dev-tools/skills/walk-pr/SKILL.md
claude-plugins/manifest-dev/skills/auto/SKILL.md
claude-plugins/manifest-dev/skills/check-pr/SKILL.md
claude-plugins/manifest-dev/skills/define/SKILL.md
claude-plugins/manifest-dev/skills/design/SKILL.md
claude-plugins/manifest-dev/skills/do/SKILL.md
claude-plugins/manifest-dev/skills/done/SKILL.md
claude-plugins/manifest-dev/skills/escalate/SKILL.md
claude-plugins/manifest-dev/skills/figure-out-team/SKILL.md
claude-plugins/manifest-dev/skills/figure-out/SKILL.md
claude-plugins/manifest-dev/skills/init-context/SKILL.md
claude-plugins/manifest-dev/skills/just-auto/SKILL.md
claude-plugins/manifest-dev/skills/just-do/SKILL.md
claude-plugins/manifest-dev/skills/next-ticket/SKILL.md
claude-plugins/manifest-dev/skills/poll-slack/SKILL.md
claude-plugins/manifest-dev/skills/review-design/SKILL.md
claude-plugins/manifest-dev/skills/review-writing/SKILL.md
claude-plugins/manifest-dev/skills/run-ticket/SKILL.md
claude-plugins/manifest-dev/skills/sweep-tickets/SKILL.md
claude-plugins/manifest-dev/skills/ticket-up/SKILL.md
claude-plugins/PLUGIN_TEMPLATE/skills/example/SKILL.md
claude-plugins/manifest-dev-tools/skills/teach-me/SKILL.md
claude-plugins/manifest-dev/skills/chat-surface/SKILL.md
claude-plugins/manifest-dev/skills/review-code/SKILL.md

Metadata

Files
0
Version
d76241f
Hash
a4279ffa
Indexed
2026-09-22 22:25

ホーム - Wiki
Copyright © 2011-2026 iteam. Current version is 2.155.2. UTC+08:00, 2026-09-29 12:51
浙ICP备14020137号-1