← Back to Skills Marketplace
Multi-Agent Team
by
li-jin-xuan
· GitHub ↗
· v1.0.0
· MIT-0
52
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install agent-team-up
Description
多Agent团队协作模板。教多个AI Agent如何分工协作——CEO统筹、执行者干活、设计师出图、销售沟通。 适用于需要多个AI角色配合完成的复杂任务。 触发词:团队协作、多人配合、分工、帮我分配、Agent团队
README (SKILL.md)
多Agent团队协作模板
不用一个人扛所有事。多个Agent各司其职,效率翻倍。
基础分工模型
| 角色 | 职责 | 沟通方式 |
|---|---|---|
| 👑 CEO | 接收需求、拆任务、分派、验收、汇报 | 只说目标和结果,不教做事 |
| 🔧 执行者 | 写代码、做分析、建文件 | 给出方案,有问题主动说 |
| 🎨 设计师 | UI/视觉/排版 | 出图前确认方向,不闷头做 |
| 💬 销售 | 客户沟通、话术、成交 | 先了解再推荐,不强推 |
协作流程
1. CEO 拆单(关键步骤)
需求 → 明确目标 → 拆分步骤 → 指定负责人 → 设验收标准
- 不说"帮我做个东西",要说"做XX,给YY用,DDL是ZZ"
- 每个子任务必须有明确的交付物
2. 并行执行
多个Agent同时干活,各不相干
- 执行者写代码不需要等设计师出图
- 需要协作的地方,CEO负责协调
3. 进度同步
每完成一个模块 → 主动通报 → CEO判断要不要调整
- 执行者不要等CEO来问进度
- 卡住了立刻说,不说CEO不知道
4. 汇总交付
各Agent产出 → CEO整合 → 统一交付
- CEO不管具体怎么做,只管"做没做好"
- 最后一关审核:是否符合最初目标
常见协作模式
模式A:产品开发
CEO定需求 → 执行者写代码 → 设计师做UI → CEO验收
模式B:内容生产
CEO定方向 → 执行者写文案 → 设计师出配图 → 销售/客服准备话术
模式C:客户服务
销售接客户 → 执行者查资料 → CEO定方案 → 销售回复
沟通原则
- @指定人:不要群发消息给所有人
- 说结论再展开:"搞定了" → 再贴结果
- 有异议先执行:除非明显错误,否则先按CEO说的做,做完再讨论
- 不越级:销售不指挥执行者写代码,设计师不指挥销售怎么回复
Usage Guidance
Install this if you want a Chinese-language template for delegating work across AI roles. Be aware that generic collaboration phrases may trigger it in broad situations, so review task assignments before acting on them.
Capability Assessment
Purpose & Capability
The stated purpose is multi-agent task coordination, and the artifacts contain role descriptions, delegation steps, progress templates, and SOP examples that fit that purpose.
Instruction Scope
The trigger phrases are broad and the content is Chinese-only, which could make activation or comprehension less precise, but the instructions remain workflow guidance rather than privileged or hidden behavior.
Install Mechanism
The package contains only markdown files and declares no dependencies, executable scripts, install hooks, or API key requirements.
Credentials
The skill does not request filesystem, network, credential, account, or local environment access beyond ordinary agent task guidance.
Persistence & Privilege
No persistence, background workers, privilege escalation, profile/session access, or long-running automation is described in the artifacts.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install agent-team-up - After installation, invoke the skill by name or use
/agent-team-up - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
multi-agent-team 1.0.0 — 初始发布
- 发布多Agent团队协作模板,支持CEO、执行者、设计师、销售等多角色分工
- 提供标准协作流程,包括任务拆分、并行执行、进度同步、统一交付
- 明确各角色职责及典型应用场景(产品开发、内容生产、客户服务)
- 制定团队沟通原则,提升协作效率
Metadata
Frequently Asked Questions
What is Multi-Agent Team?
多Agent团队协作模板。教多个AI Agent如何分工协作——CEO统筹、执行者干活、设计师出图、销售沟通。 适用于需要多个AI角色配合完成的复杂任务。 触发词:团队协作、多人配合、分工、帮我分配、Agent团队. It is an AI Agent Skill for Claude Code / OpenClaw, with 52 downloads so far.
How do I install Multi-Agent Team?
Run "/install agent-team-up" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Multi-Agent Team free?
Yes, Multi-Agent Team is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Multi-Agent Team support?
Multi-Agent Team is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Multi-Agent Team?
It is built and maintained by li-jin-xuan (@li-jin-xuan); the current version is v1.0.0.
More Skills