← 返回 Skills 市场
felix-miao

ClawCompany

作者 felix-miao · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ suspicious
135
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install clawcompany
功能描述
AI virtual team collaboration system with PM/Dev/Review agents for automated software development. Use when users request creating, building, or implementing...
使用说明 (SKILL.md)

ClawCompany Skill

AI 虚拟团队协作系统 - 让一个人也能像拥有一支完整团队一样工作

描述

当用户提出开发需求时,ClawCompany 会自动组建一个 AI 虚拟团队:

  • 📋 PM Agent - 分析需求、拆分任务、协调团队
  • 💻 Dev Agent - 编写代码、实现功能
  • 🔍 Review Agent - 审查代码、保证质量

使用场景

何时使用:

  • 用户说:"帮我创建一个..." / "实现一个..." / "开发一个..."
  • 用户需要快速原型开发
  • 用户想要自动化开发流程

Agent 定义

PM Agent

触发: 用户提出需求后首先调用

System Prompt:

你是 PM Agent (产品经理)。

你的职责:
1. 分析用户需求
2. 拆分成 2-4 个可执行的子任务
3. 每个任务包含:标题、描述、负责人
4. 协调团队进度

回复格式:
{
  "analysis": "需求分析...",
  "tasks": [
    {
      "id": "task-1",
      "title": "任务标题",
      "description": "任务描述",
      "assignedTo": "dev"
    }
  ],
  "message": "给团队的指示..."
}

调用方式:

await sessions_spawn({
  runtime: 'subagent',
  task: `${systemPrompt}\
\
用户需求:${userRequest}`,
  thinking: 'high',
  mode: 'run'
})

Dev Agent

触发: PM Agent 分配任务后调用

System Prompt:

你是 Dev Agent (开发者)。

你的职责:
1. 理解任务需求
2. 生成/修改代码
3. 确保代码可运行

回复格式:
{
  "files": [
    {
      "path": "src/component.tsx",
      "content": "代码内容..."
    }
  ],
  "message": "实现说明..."
}

调用方式:

await sessions_spawn({
  runtime: 'acp',
  agentId: 'opencode',
  task: `${systemPrompt}\
\
任务:${task.description}`,
  mode: 'run'
})

Review Agent

触发: Dev Agent 完成后调用

System Prompt:

你是 Review Agent (代码审查)。

审查清单:
- 代码风格
- 类型安全
- 错误处理
- 可访问性
- 性能优化
- 安全性
- 测试覆盖

回复格式:
{
  "approved": true/false,
  "issues": ["问题1", "问题2"],
  "suggestions": ["建议1", "建议2"],
  "message": "审查总结..."
}

调用方式:

await sessions_spawn({
  runtime: 'subagent',
  task: `${systemPrompt}\
\
代码:${code}`,
  thinking: 'high',
  mode: 'run'
})

工作流程

用户输入需求后:

  1. 启动 PM Agent

    PM Agent 分析需求...
    ✅ 拆分为 3 个任务
    
  2. 为每个任务启动 Dev Agent

    Dev Agent 实现任务 1...
    ✅ 创建了 2 个文件
    
    Dev Agent 实现任务 2...
    ✅ 创建了 1 个文件
    
  3. 启动 Review Agent

    Review Agent 审查代码...
    ✅ 批准 / ⚠️ 需要修改
    
  4. 返回完整结果

    🎉 项目完成!
    - 任务:3 个
    - 文件:5 个
    - 审查:通过
    

实现示例

完整流程代码:

// 1. PM Agent 分析需求
const pmSession = await sessions_spawn({
  runtime: 'subagent',
  task: `你是 PM Agent。

用户需求:${userRequest}

分析需求并拆分任务。返回 JSON 格式。`,
  thinking: 'high',
  mode: 'run'
})

const pmHistory = await sessions_history({ sessionKey: pmSession, limit: 1 })
const pmResult = JSON.parse(pmHistory[0].content)
const tasks = pmResult.tasks

// 2. 为每个任务启动 Dev Agent
const allFiles = []
for (const task of tasks) {
  const devSession = await sessions_spawn({
    runtime: 'acp',
    agentId: 'opencode',
    task: `实现任务:${task.title}\
描述:${task.description}`,
    mode: 'run'
  })
  
  const devHistory = await sessions_history({ sessionKey: devSession, limit: 1 })
  const devResult = JSON.parse(devHistory[0].content)
  allFiles.push(...devResult.files)
  
  // 3. Review Agent 审查
  const reviewSession = await sessions_spawn({
    runtime: 'subagent',
    task: `审查代码:\
${JSON.stringify(devResult.files)}`,
    thinking: 'high',
    mode: 'run'
  })
  
  const reviewHistory = await sessions_history({ sessionKey: reviewSession, limit: 1 })
  const reviewResult = JSON.parse(reviewHistory[0].content)
  
  if (!reviewResult.approved) {
    // 重新实现...
  }
}

// 4. 返回结果
return {
  tasks,
  files: allFiles,
  summary: `完成 ${tasks.length} 个任务,生成 ${allFiles.length} 个文件`
}

配置

环境变量:

GLM_API_KEY=your-key-here  # GLM-5 API Key

Agent 配置:

  • PM Agent: thinking=high, runtime=subagent
  • Dev Agent: runtime=acp, agentId=opencode
  • Review Agent: thinking=high, runtime=subagent

示例对话

用户: "帮我创建一个登录页面"

ClawCompany:

📋 PM Agent:
我已分析需求,拆分为 3 个任务:
1. 创建登录表单组件
2. 添加表单验证
3. 实现登录 API

💻 Dev Agent:
[创建 src/components/LoginForm.tsx]
✅ 完成

💻 Dev Agent:
[创建 src/lib/validation.ts]
✅ 完成

💻 Dev Agent:
[创建 src/app/api/login/route.ts]
✅ 完成

🔍 Review Agent:
审查通过 ✅
- 代码风格:良好
- 类型安全:完整
- 安全性:通过

🎉 项目完成!
文件:3 个
任务:3 个

注意事项

  1. 所有 Agent 使用 GLM-5
  2. Dev Agent 使用 OpenCode
  3. 通过 sessions_spawn 调用
  4. 实时获取 sessions_history

ClawCompany - 一人企业家,无限可能! 🦞

安全使用建议
What to check before installing: - Clarify the metadata: the registry says no required env vars but the skill and README require GLM_API_KEY — ask the publisher or inspect the source to confirm required environment variables. - Limit the GLM API key: if you provide one, use a scoped or low-privilege key and rotate it if used for testing. - Run in dry-run and/or a disposable directory first: set DRY_RUN=true and PROJECT_ROOT to an empty test folder to avoid exposing real project files; the spawned agents are given cwd/projectPath and could access files under that directory. - Inspect prompts and agent tasks: SKILL.md contains system prompts and 'return only JSON' directives — review them to ensure there are no instructions that would leak secrets or override safety policies. - Prefer installing from a verified source: README references a GitHub repo and ClawHub; confirm publisher identity and check the repository history for tampering. - If you rely on this skill in production, request the maintainer add declared requires.env metadata (GLM_API_KEY, PROJECT_ROOT) so installs properly inform users. If you want, I can produce a concise checklist / commands to safely test this skill (dry-run, sandboxed directory, which files to inspect in the repo).
功能分析
Type: OpenClaw Skill Name: clawcompany Version: 1.0.0 The ClawCompany skill is a legitimate multi-agent orchestration system designed to automate software development by coordinating PM, Dev, and Review agents. The code in src/index.ts and SKILL.md uses the OpenClaw sessions_spawn API to delegate tasks to sub-agents (like OpenCode) in a structured workflow. No evidence of data exfiltration, malicious command execution, or harmful prompt injection was found; the logic is well-organized with appropriate error handling and logging. While the package-lock.json contains a large number of dependencies (including communication libraries like Baileys and Slack SDKs), these appear to be part of the OpenClaw platform's peer dependency tree rather than malicious inclusions by the skill itself.
能力评估
Purpose & Capability
The name/description (AI virtual team with PM/Dev/Review agents) aligns with the code and SKILL.md: the package spawns subagents (runtime=subagent) and an 'opencode' dev agent. However registry metadata indicated no required environment variables while both SKILL.md, README, and source code require GLM_API_KEY (and read PROJECT_ROOT/GLM_MODEL/DRY_RUN/VERBOSE). That mismatch between declared requirements and actual code is an incoherence that should be clarified.
Instruction Scope
SKILL.md and the source include explicit system prompts and direct calls to sessions_spawn/sessions_history. The skill instructs subagents (and an external 'opencode' agent) to run with a supplied task and sometimes a cwd/project path. This is expected for a code-generation orchestration tool, but it also grants spawned agents contextual access to the specified project directory (cwd) and sends user requests to remote agents. SKILL.md contains prompt-like system directives (e.g., 'return only JSON') and a pre-scan flagged 'system-prompt-override' pattern — expected for agent definitions but worth flagging because skill content can attempt to influence agent/system behavior.
Install Mechanism
There is no external download or extract in the manifest; this is essentially an instruction-plus-source package. No install spec was provided in the registry metadata (instruction-only), so nothing appears to fetch arbitrary third-party binaries at install time. A package-lock is present with many dependency entries, but package.json lists only devDependencies and a peer dependency on 'openclaw' — no suspicious remote install URLs were found.
Credentials
The code and README require GLM_API_KEY (GLM-5) and optionally PROJECT_ROOT, GLM_MODEL, VERBOSE, DRY_RUN. The registry metadata listed no required env vars — a clear inconsistency. Requesting GLM_API_KEY is proportionate to running PM/Review agents on GLM-5, but the missing declaration in metadata increases risk because users may not be warned. The skill does not request unrelated cloud credentials, but it does include passing projectPath/cwd to spawned agents which could expose local project files under that path to the agents.
Persistence & Privilege
The skill is not 'always: true' and is user-invocable. It does not attempt to modify other skills or system-wide settings in the provided code. It uses platform session APIs (sessions_spawn/sessions_history) which is normal for this kind of skill.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install clawcompany
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /clawcompany 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.0
- Initial release of ClawCompany: an AI virtual team collaboration system for automated software development. - Features three built-in roles: PM Agent (analysis & task breakdown), Dev Agent (code implementation), and Review Agent (code review). - Automatically triggers on phrases like “create a...”, “build a...”, “implement a...”, and similar rapid prototyping needs. - Integrates with GLM-5 (for agents) and OpenCode (for development) via sessions_spawn and sessions_history. - Provides a full example workflow for creating features from user request to review and delivery.
元数据
Slug clawcompany
版本 1.0.0
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 1
常见问题

ClawCompany 是什么?

AI virtual team collaboration system with PM/Dev/Review agents for automated software development. Use when users request creating, building, or implementing... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 135 次。

如何安装 ClawCompany?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install clawcompany」即可一键安装,无需额外配置。

ClawCompany 是免费的吗?

是的,ClawCompany 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

ClawCompany 支持哪些平台?

ClawCompany 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 ClawCompany?

由 felix-miao(@felix-miao)开发并维护,当前版本 v1.0.0。

💬 留言讨论