بنقرة واحدة
remove-teammate
让某个 teammate 下岗:优雅完成交接、归档其产出、从 team 中移除。 Team lead 可自主调用(用户自然语言同意后)。只在 team lead 上下文有效。
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
القائمة
让某个 teammate 下岗:优雅完成交接、归档其产出、从 team 中移除。 Team lead 可自主调用(用户自然语言同意后)。只在 team lead 上下文有效。
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
| name | remove-teammate |
| description | 让某个 teammate 下岗:优雅完成交接、归档其产出、从 team 中移除。 Team lead 可自主调用(用户自然语言同意后)。只在 team lead 上下文有效。 |
| disable-model-invocation | false |
| allowed-tools | Read Write Edit Glob Grep Bash |
/remove-teammate — 让某个 teammate 下岗先从对话 context 推断:你应该已经知道自己是 team lead。无法推断时按 /evaluate-team 的检查逻辑。若不是 team lead → 立即停止并警告。
从用户描述或 /evaluate-team 的结果中确定要移除的 teammate 昵称。
向用户确认:
你想让 <Teammate 昵称> 下岗吗?
它的近期产出:
- <列出 roundtable 中 from 该 teammate 的文件清单>
它的手头任务:
- <列出 IN_PROGRESS 或 OPEN 且 to 该 teammate 的任务>
移除前,我需要:
1. **要求它做最终 checkpoint**(它自己调 /checkpoint 写 working-context.md 和 completed.md)
2. 把它手头的任务交接给其他 teammate 或放回待分配
3. 把它的产出(DONE 报告、代码等)收尾归档
4. 从 Claude Code agent-team 中正式移除它
5. 更新 TEAMMATE_INFO.json:把它从 active_teammates 移到 offboarded_teammates
6. 更新 team_recipes/ 的最新 recipe,追加 Amendment 记录此次下岗
你确认移除吗?
等用户明确同意。
关键:最终 checkpoint 必须由被移除的 teammate 自己完成(rule #1 低耦合,lead 不能替它写工位文件)。
若该 teammate 当前
status=benched(已临时下线、不在线):跳过下面的 SendMessage——它在 bench 时已做过最终 checkpoint,直接沿用其工位上的 working-context.md,进入 Phase 2。(benched → offboarded 是合法转移。)
通过 SendMessage 给该 teammate:
Before shutdown, run /checkpoint to persist your final state. This is your final
checkpoint — after it, I will remove you from the team. Your workstation files
(working-context.md, completed.md, commitments.md) will be preserved at
_agent_team_work_zone/<SELF>_team/teammates/<昵称>/ for audit and potential future
reactivation under a different task.
等待 teammate 确认 "Checkpoint written." 再进入 Phase 2。
如果 teammate 因为任何原因(卡死 / session 已死 / 无响应)无法做 final checkpoint:
扫 _agent_team_work_zone/<SELF>_team/roundtable/ 中:
from: <SELF>_team/<teammate> 且 status: IN_PROGRESS 的文件(它自己提交的但未完成)to: <SELF>_team/<teammate> 且 status: OPEN 或 IN_PROGRESS 的文件(别人派给它还没做完)对每个未完成项向用户询问:
未完成任务交接:
1. <file1.md> (IN_PROGRESS, from: <teammate>)
进度: <摘要>
选项:
(a) 让它先收尾后再下岗
(b) 强制接管 — 由 lead 或另一 teammate 继续
(c) 放弃这项工作
2. <file2.md> (OPEN, to: <teammate>)
选项:
(a) 重派给另一 teammate (指定谁)
(b) 放回待分配,记录到 <SELF>_team/TODO.md
(c) 取消
你的决定?
按用户决定:
to 字段<SELF>_team/TODO.md 追加一项,并把原文件归档或标为 WITHDRAWN/remove-teammate把该 teammate 的 RESOLVED 文件(from: <teammate> 且 to 指向 lead 或其他已离场 teammate 的)从 roundtable 移到 <SELF>_team/archive/:
mv _agent_team_work_zone/<SELF>_team/roundtable/<file>.md _agent_team_work_zone/<SELF>_team/archive/
守则 #8 依然适用:
from: <SELF>_team/<teammate> 的文件 lead 有权归档(因为 teammate 即将下岗,相当于 lead 接管它的发文权)to: ALL 的文件不归档cc 中的文件不归档产出一段 natural-language remove 指令(Claude Code 内置机制识别):
从当前 <SELF> team 中移除 teammate <昵称>。它的未完成任务已经交接(详见 roundtable),
它的产出已归档到 archive/。请正式 remove 它。
路径:_agent_team_work_zone/<SELF>_team/TEAMMATE_INFO.json
把被移除的 teammate 从 active_teammates 数组移除,并 append 到 offboarded_teammates 数组:
{
"name": "<被移除的昵称>",
"offboarded_at": "<ISO8601 当前时间>",
"reason": "<一句话原因:task completed / redundant role / stuck / user decision>"
}
同时更新顶层 updated_at。
jq 示例(如果可用):
name="<昵称>"
reason="<原因>"
ts="$(date -u +%Y-%m-%dT%H:%M:%SZ)"
jq --arg n "$name" --arg r "$reason" --arg ts "$ts" '
.offboarded_teammates += [{name: $n, offboarded_at: $ts, reason: $r}] |
.active_teammates |= map(select(.name != $n)) |
.updated_at = $ts
' _agent_team_work_zone/<SELF>_team/TEAMMATE_INFO.json > /tmp/info.json && \
mv /tmp/info.json _agent_team_work_zone/<SELF>_team/TEAMMATE_INFO.json
不删除 teammate 的工位目录 _agent_team_work_zone/<SELF>_team/teammates/<昵称>/——保留原位作为历史审计(Rule 1 低耦合 + 便于日后如果相同任务重启可参考历史 working-context.md)。
找到 <SELF>_team/team_recipes/ 下最新的 recipe,在末尾追加:
---
## Amendment — YYYY-MM-DD HH:MM — remove-teammate
### 下岗
- **昵称**: <teammate>
- **原因**: <来自用户或 /evaluate-team 的判断>
### 交接处理
- 未完成任务交接: <摘要>
- 产出归档: N 份 RESOLVED 移到 archive/
### 当前 team 人数: <剩余人数>
✅ <Teammate 昵称> 已下岗
交接:
- <file1.md> → 重派给 <new teammate>
- <file2.md> → 放回 TODO
- ...
归档:
- <N 份文件> → <SELF>_team/archive/
Team 更新:
- 人数: <before> → <after>
- TEAMMATE_INFO.json: active_teammates -1, offboarded_teammates +1
- team_recipes/<latest>.md 已追加 Amendment
- Claude Code agent-team 已正式 remove
- 工位目录保留在 teammates/<昵称>/(审计用,不删除)
当前 team 剩余成员:
- <list>
teammates/<昵称>/ 作为历史审计 + 潜在的 reactivation 参考to: ALL、cc 中的文件绝不归档Team lead invokes after session restart: reads TEAMMATE_INFO.json to rebuild the team by using the Agent tool to spawn fresh sessions for each active teammate (name must be <slug>-<role>; permission mode inherits the lead, not set per-teammate at spawn), guiding each to read its workstation's working-context.md (Part A snapshot + Part B work journal) to self-recover state. No-arg invocation wakes only active/idle and **skips benched** (temporarily offline); `/reactivate-team <name>` wakes back one specified (usually benched) teammate. This is the ONLY way to "recover a team across sessions" — Claude Code does not auto-respawn teammates.
Structured flow for a team lead to assemble a Claude Code agent team. 6 phases: task decomposition → lineup proposal (selected from role_archetypes) → plan-mode gating → adversarial check → user confirmation → spawn each teammate via the Agent tool (name must be <slug>-<role>; permission mode inherits the lead, not set per-teammate at spawn) + save team_recipe. After Phase 5 user confirmation, Phase 6 spawns immediately with no additional confirmation. The agent can invoke autonomously; if currently a flat workstation, it will first guide through /promote-to-team.
Team lead 在 session 重启后调用,读 TEAMMATE_INFO.json 重建 team: 为每个 active teammate 用 Agent 工具 spawn 新 session(name 必须为 <slug>-<role>;权限模式继承 lead,不在 spawn 时单设), 引导它读自己工位的 working-context.md(Part A 快照 + Part B 工作日志)恢复状态。 无参调用只唤醒 active/idle,**跳过 benched**(临时下线);`/reactivate-team <name>` 单独唤回指定的(通常是 benched)teammate。这是"跨 session 团队恢复"的唯一方式—— Claude Code 不会自动 respawn teammate。
Team lead 组建 Claude Code agent team 的结构化流程。6 阶段:任务分解 → 阵容提案 (从 role_archetypes 选) → plan-mode gating → 对抗性检查 → 用户确认 → 用 Agent 工具(name 必须为 <slug>-<role>;权限模式继承 lead,不在 spawn 时单设)逐一 spawn 每个 teammate + 保存 team_recipe。Phase 5 用户确认后直接 spawn,不再额外确认。Agent 可自主调用;若当前是扁平工位,会先引导 /promote-to-team。
Add a teammate to an existing team: runs a simplified spawn-team flow (single person only). Team lead can invoke autonomously (after natural-language user consent). Valid only in team lead context.
Temporarily take a teammate offline (benched): have it write a final checkpoint, close its session to free an online slot, but **keep its full record + workstation + docs**, and set its status to benched in TEAMMATE_INFO.json. Unlike /remove-teammate (permanent offboard), benched means "will come back" — wake it later with `/reactivate-team <name>`. Team lead invokes autonomously (after user consent). Only valid in team lead context.