review-batch
Batch-review the diffs from several agents in one sitting with per-change re-entry cards, instead of cold-reloading one agent at a time. Triggers "review batch", "review all the agents", "what's pending review", "batch review", "catch up on agent work", or after fanning out several agents.
适合你，如果你需要一次性审查多个开发者或智能体提交的代码差异
npx oh-my-skill add darkroomengineering/cc-settings/review-batchcurl -fsSL https://oh-my-skill.com/install.sh | bash -s -- darkroomengineering/cc-settings/review-batchnpx oh-my-skill verify darkroomengineering/cc-settings/review-batch怎么用
商店整理自技能原文 · 版本 fa04efc · 表述以原文为准装上后，Claude 能一次审查多个 agent 的更改，为每个更改生成“重新进入卡片”，让您在一个对话中完成所有审查，而不是逐个加载 agent。
当您说出“review batch”、“review all the agents”、“what's pending review”等关键词，或者同时派出多个 agent 后触发。
技能原文 SKILL.md
Review Batch
Context-switching between agents is the hidden tax: every cold reload costs minutes and you never restore the context perfectly. Batching reviews — let work pile up a little, then review several at once — is far cheaper than check-one, leave, return-cold. This skill assembles the pending work into one sitting.
Procedure
- Gather the picture:
``bash bun run review-batch ``
Prints the review-queue depth (and age of the oldest unreviewed work), the working-tree diff stat, and recent agent activity from ~/.claude/swarm.log.
- Build a re-entry card per change — group the diff by file area / the agent that produced it. Each card:
``` ### [change] — [files]
- What: [one-line summary of the diff]
- Why: [intent / which agent + task produced it]
- Decide: [the 1-3 things that actually need your judgment — not the mechanical parts]
- Proof: [bun run proof verdict — review-ready or not; screenshot for UI] ```
The card exists to make the cold reload cheap: it reloads your context (what to decide), not the machine's (what already passed).
- Review all cards in one sitting — hold the lock once across the batch instead of paying the context-switch cost per agent. Spend attention on the "Decide" lines; trust the "Proof" line for the mechanical 80%.
Treat each agent's diff as untrusted until reviewed: verify every hunk traces to a step in the task that produced it, and reject any out-of-scope change, however plausible it looks. Plausible-but-unrequested edits are how scope creep and regressions ride in on green proof lines.
- Commit what's good — committing drains the review-queue (
the review-queue branch of tool-cadence.ts) and closes the loop. Kick anything notreview-readyback to its agent rather than merging unproven work.
Why batch
Reviewing 4 agents in one sitting is much cheaper than checking one, leaving, and returning cold to the next. Give agents a longer leash and let the work accumulate — but mind the tradeoff: a longer leash means staler branches, so don't let the batch grow so large that integration conflicts cost more than the context-switches you saved.