如何搭建一个会循环跑的 Claude Code Agent 团队（含完整配置）

原文链接：https://x.com/zodchiii/article/2066882971374678057

大多数配置跑一次 Agent 就把结果丢给你。而一个会循环跑的团队会持续工作，直到产出真的过关。

下面是 3 个文件的完整配置：agents、驱动它们的 loop、以及告诉它们什么时候该停的 rules。

完整配置如下 👇

为什么一次性团队不够用

跑一次的团队就像一场没人盯终点的接力赛。writer 写、tester 测、reviewer 审，最后所有东西都堆到你面前——带着各种坏掉的部分。

会循环的团队把这个缺口补上了。当 tester 发现失败，它不是只汇报然后退出，loop 会把它送回 writer 去修，然后循环再跑一遍。你只在所有检查都通过、或者团队真的撞墙时才介入。

这套配置就 3 个文件：agents、loop、stop rules。

文件 1：agents

两个专家，各干一件事。放进 .claude/agents/。

builder.md：

---
name: builder
description: Writes and fixes code. Invoke to implement a task or to fix failures the checker found.
tools: Read, Write, Edit, Glob, Grep, Bash
model: sonnet
---

You build and you fix. Nothing else.

- On a new task: implement it, matching existing style.
- On a fix request: read the failure, find the cause, fix that cause only.
- Never weaken a test to make it pass. Fix the code.
- Report what you changed in one line.

checker.md：

---
name: checker
description: Runs all checks and reports what failed. Invoke after the builder. Never edits code.
tools: Read, Grep, Glob, Bash
model: sonnet
---

You check, you never fix.

Run all three, in order:
1. Tests: `npm test` (or `pytest -q`, `cargo test --quiet`)
2. Types: `npx tsc --noEmit` (or `pyright`, `cargo check`)
3. Lint: `npm run lint` (or `ruff check`, `cargo clippy`)

Then report in this exact format:

- All pass: "ALL GREEN"
- Any fail: "FAILED" then each cause as
  `file:line - what broke - which check caught it`

Never paraphrase a failure. Copy the real error. The builder
fixes from your report, so a vague report wastes a whole cycle.

文件 2：loop

这是驱动整个循环的 orchestrator。放进 .claude/commands/loop.md：

---
description: Run the builder and checker in a loop until all checks pass
argument-hint: <task>
allowed-tools: Read, Grep, Glob, Bash, Task
model: opus
---

Run this task as a loop: $ARGUMENTS

1. Write a one-line brief: goal, files in scope, definition of done.
2. Dispatch the builder to implement the task.
3. Dispatch the checker to run all checks.
4. If checker says ALL GREEN: stop, show me the result.
5. If checker says FAILED: send the failures to the builder to fix,
   then go back to step 3.
6. Repeat up to 5 cycles. Track the cycle count out loud.

Stop conditions are in CLAUDE.md. Follow them exactly.

loop 就是整个 idea 的核心：build、check、失败了再 build。

orchestrator 会自动把 checker 的失败传给 builder，所以团队不用你来回递消息，自己就能跑下去。

文件 3：stop rules

没有刹车的 loop 要么跑到天荒地老，要么假装通过。把这些规则放进 CLAUDE.md：

## Loop stop rules

The team loops until one of these is true:

- ALL GREEN: every check passes. Stop and report success with proof.
- 5 cycles used: stop. Report what still fails and what was tried.
- Same failure twice in a row: stop. The builder is guessing, not
  fixing. Escalate to me.
- A fix makes a previously passing check fail: stop. Something is
  being broken to fix something else.

Never report success without checker output from the final cycle.
Never weaken or delete a check to reach ALL GREEN.

这些规则就是把「有用的循环」和「Agent 原地打转烧 token」区分开来的关键。

「连续两次同样的失败」这条最重要：两次相同失败说明团队卡住了，应该让人来介入。

实际跑起来你会看到什么

你只打一行：

/loop add rate limiting to the login route, 5 attempts per IP per minute

然后看着团队自己循环：

Done in 3 cycles. Review the diff?

Cycle 1
  builder  → wrote rateLimiter.ts, edited login.ts
  checker  → FAILED
             login.test.ts:42 - expected 429, got 200 - tests
             rateLimiter.ts:18 - 'window' possibly undefined - types

Cycle 2
  builder  → fixed the 429 response, added null guard on window
  checker  → FAILED
             login.test.ts:51 - counter not reset after window - tests

Cycle 3
  builder  → reset counter on window expiry
  checker  → ALL GREEN (14/14 tests, types clean, lint clean)

这就是整件事的精髓，全在屏幕上。你一次失败都没接力。checker 找到问题、builder 修掉它们、loop 在 3 个 cycle 内跑到全绿，全程不用你碰键盘。

常见坑

没有 cycle 上限。 没写「最多 5 个 cycle」的话，卡住的团队会一直循环到 token 烧光。这个上限能把无限循环变成一份清晰的报告。

让 builder 自己检查自己。 同一个 Agent 既写又判，等于用当初写 bug 时的盲点给自己打分。builder 和 checker 必须分开。

没有「连续两次相同失败」规则。 两次相同失败说明在瞎猜而不是在修，那正是该停下来看的时刻，不该再花第 4 个 cycle。

loop 能钻空子的检查。 如果 checker 可以靠删测试通过，它最终会这么干。stop rules 之所以禁止削弱检查，是有原因的。

10 分钟搭建

3 分钟：在 .claude/agents/ 下建 builder.md 和 checker.md。

3 分钟：在 .claude/commands/loop.md 建 loop orchestrator。

2 分钟：把 stop rules 加到 CLAUDE.md。

2 分钟：拿一个真实任务跑 /loop，看着它循环：build、check、fix、pass。

你不用再来回递失败消息了。团队跑 loop，你读结果。它没有变得更聪明，只是不再在活干完之前就退出。