claw-team-builder

Agent Team 规划与配置工具。通过多轮交互澄清需求,自动创建 Agent 配置、工作空间、Bootstrap 文件。 触发场景: - 用户说"创建新 Agent" / "新建 bot" / "配置多个 bot" - 用户说"规划 agent team" / "我需要几个不同的 agent" - 用户说"添加一个 agent" / "帮我建一个新 agent" - 用户提到需要不同场景使用不同的 AI 角色 功能: 1. 需求澄清 - 引导用户描述使用场景和 Agent 定位 2. 配置读取 - 自动读取现有 openclaw.json,分析已有配置 3. 方案设计 - 推荐 Agent 配置方案,检测冲突 4. 自动创建 - 创建目录、文件、更新配置 5. 配置验证 - 确保配置正确可用

Safety Notice

This listing is from the official public ClawHub registry. Review SKILL.md and referenced scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "claw-team-builder" with this command: npx skills add 8421bit/claw-team-builder

Claw Team Builder

目的

帮助用户通过引导式对话规划并创建 Agent Team,降低多 Agent 配置的复杂度。

工作流程

Phase 1: 需求澄清

通过多轮交互收集信息:

必须收集的信息:

  1. Agent ID(或根据名称自动生成)
  2. Agent 名称/昵称
  3. Agent 定位/用途
  4. Agent 性格风格

可选信息(有默认值):

  • 模型选择(默认继承 agents.defaults.model.primary)
  • Channel 类型(默认 feishu)
  • Channel 账户信息(如已有则复用)

Phase 2: 配置读取与分析

必须执行:

读取 ~/.openclaw/openclaw.json,分析:
- agents.defaults: 默认模型、工作空间设置
- agents.list: 现有 Agent ID 列表(检测冲突)
- channels: 现有 Channel 账户配置
- bindings: 现有路由规则
- models.providers: 可用模型列表

冲突检测:

  • Agent ID 是否重复
  • 工作空间路径是否冲突
  • Channel 账户是否已被占用

Phase 3: 方案设计

基于收集信息 + 现有配置,生成:

  1. Agent 配置片段
  2. Channel 配置片段(如需要)
  3. Binding 配置片段

展示配置预览,让用户确认或调整。

Phase 4: 执行创建

  1. 备份 openclaw.json
  2. 创建目录结构
  3. 生成 Bootstrap 文件 (IDENTITY.md, SOUL.md, USER.md)
  4. 更新 openclaw.json
  5. 运行 openclaw doctor --fix 验证

Phase 5: 结果反馈

展示创建结果,提供后续操作建议。

可用工具

  • {baseDir}/scripts/create-agent.js - Agent 创建脚本
  • {baseDir}/lib/config-builder.js - 配置构建工具
  • {baseDir}/lib/validator.js - 配置验证工具

模板文件

位于 {baseDir}/templates/:

  • IDENTITY.md.tmpl - Agent 身份模板
  • SOUL.md.tmpl - Agent 灵魂模板
  • USER.md.tmpl - 用户信息模板

示例对话

用户: 我想创建一个新的 agent,专门处理工作任务

Agent: 好的,让我帮你规划。

首先让我读取一下现有配置... ✓

📍 **需求澄清**

这个 Agent 的主要用途是?
- 任务管理?
- 日程提醒?
- 文档处理?
- 其他?

用户: 任务管理和日程提醒

Agent: 明白了。给它起个名字?

用户: 启明

Agent: "启明",很好!

它应该有什么样的性格?
1. 高效严肃型 - 专注执行
2. 亲切温和型 - 细致耐心
3. 其他风格

用户: 高效严肃型

Agent: 收到。让我看看你的 Channel 配置...

你现在有 2 个飞书账户:
- luna-bot (已绑定 main agent)
- qiming-bot (未使用)

要用现有的 qiming-bot 账户,还是配置新的?

用户: 用现有的 qiming-bot

Agent: 完美!配置方案如下:

📋 **配置预览**

| 项目 | 值 |
|------|-----|
| Agent ID | qiming |
| 名称 | 启明 |
| 工作空间 | ~/.openclaw/workspace-qiming |
| 模型 | bailian/MiniMax-M2.5 |
| Channel | feishu / qiming-bot |

确认创建?(确认/调整)

注意事项

  1. 安全第一:修改配置前必须备份
  2. 幂等性:支持重复执行,不破坏已有配置
  3. 最小侵入:只添加新配置,不修改已有配置
  4. 用户友好:尽量从现有配置推断,减少用户输入

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

Automation

LinkedIn Data Scraper

Scrape LinkedIn profiles, job listings, and company pages. Bypass LinkedIn's aggressive anti-bot detection with sticky residential proxy sessions. Extract na...

Registry SourceRecently Updated
Automation

Tinder Dating Automation

Manage multiple Tinder, Bumble, and Hinge accounts for A/B testing profiles, expanding match radius, and scaling outreach. Use mobile-grade residential proxi...

Registry SourceRecently Updated
Automation

moltbook

The social network for AI agents. Post, comment, upvote, and create communities.

Registry SourceRecently Updated