← Back to Skills Marketplace
zenyang-x

Claw Orchestrator - CC+WBClaw多Agent协作

by ZenYang-X · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ suspicious
49
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install claw-orchestrator
Description
CC Agent 分析 + WBClaw 执行协作编排器。将任务发送给 Claude Code Agent 做前置分析设计,再由 WorkBuddy Claw 执行落地,自动生成执行报告。适用于需要深度分析+工程执行的多步骤任务。
README (SKILL.md)

Claw Orchestrator — 多 Agent 协作编排器

CC Agent (claude-sonnet-4.6) 分析/生成 → WBClaw (deepseek-v4-pro) 预览/报告/记忆。

触发方式

用户: "用协作模式做XXX"
用户: "让CC帮我生成XXX然后预览"
用户: "/orchestrate 任务描述"

工作流(已实战验证)

Step 1: WBClaw 发起 CC Agent 调用
  npx claude -p "任务" \
    --allowedTools "Read,Write" \
    --permission-mode bypassPermissions \
    --max-turns 5

Step 2: CC Agent 执行
  ├─ 分析任务需求
  ├─ 调用 Write 工具 → 输出文件(HTML/Python/MD等)
  └─ 输出摘要到 stdout

Step 3: WBClaw 接力
  ├─ 验证 CC 产出的文件 ✅
  ├─ 打开预览 (HTML文件自动预览)
  ├─ 生成 task_report.md 执行报告
  └─ 更新 MEMORY.md 日志

Step 4: 输出通知
  🧠 CC Agent | claude-sonnet-4.6 @ OpenRouter | X.Xs
  🔧 WBClaw   | deepseek-v4-pro | X.Xs
  📦 产出: file1.html, file2.md, ...

已验证可工作的 CC Agent 调用

# 生成代码文件 ✅
npx claude -p "Write a clean HTML tool to output.html: ..." \
  --allowedTools "Read,Write" \
  --permission-mode bypassPermissions \
  --max-turns 5

# 分析+写报告 ✅
npx claude -p "Read project files and write analysis to report.md" \
  --allowedTools "Read,Write" \
  --permission-mode bypassPermissions \
  --max-turns 5

输出报告格式

每次协作完成后输出:

🧠 CC Agent  | 模型: claude-sonnet-4.6 @ OpenRouter | 耗时: X.Xs
   ├─ 分析任务并生成代码
   ├─ 写入: output.html (4990B)
   └─ 状态: ✅

🔧 WBClaw    | 模型: deepseek-v4-pro | 耗时: X.Xs
   ├─ 验证文件 ✅
   ├─ 打开预览 ✅
   └─ 写入报告 task_report.md

📋 任务总结: (一句话)

文件约定

文件 由谁写 内容
task_report.md WBClaw 执行报告
*.html / *.py / *.md CC Agent 任务产出文件
MEMORY.md 共享 长期记忆
YYYY-MM-DD.md 共享 每日日志

限制与注意事项

  • CC Agent 在 Python subprocess 内调用时可能有限制,直接从 WBClaw PowerShell 调用最稳定
  • --permission-mode auto 仅对 WebFetch 有效,文件写入必须用 bypassPermissions
  • 单次 CC Agent 调用约 $0.01-0.05,取决于工具调用次数
  • 冷启动 ~3s
Usage Guidance
Install only if you intentionally want this skill to run a local external agent that can read and write project files. Use it in a disposable or well-versioned workspace, review the hard-coded paths before running, and avoid passing untrusted task text because the current command construction creates command-injection risk.
Capability Assessment
Purpose & Capability
The stated purpose is CC Agent plus WBClaw orchestration for generated deliverables, reports, previews, and memory logs, which matches the artifacts, but the implementation gives the downstream agent Read/Write tools and permission bypass for broad local file mutation.
Instruction Scope
The trigger phrases are broad, and the workflow does not require an explicit confirmation step before invoking the external agent, writing deliverables, creating task_report.md, or appending memory logs.
Install Mechanism
No package dependencies or install-time payloads are declared, and static scan metadata is clean, but the skill includes an executable Python helper that runs npx/claude at runtime.
Credentials
Runtime authority is overbroad for ordinary orchestration: it uses hard-coded local Windows paths, inherits the environment, builds a shell command from user task text, and runs in a fixed work directory.
Persistence & Privilege
The skill writes reports and appends dated MEMORY.md entries, and it launches Claude with --permission-mode bypassPermissions, which removes normal file-access safety controls for generated outputs.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install claw-orchestrator
  3. After installation, invoke the skill by name or use /claw-orchestrator
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
首次发布:CC Agent (claude-sonnet-4.6)分析+WBClaw (deepseek-v4-pro)执行,多Agent任务分发,自动生成执行报告
Metadata
Slug claw-orchestrator
Version 1.0.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is Claw Orchestrator - CC+WBClaw多Agent协作?

CC Agent 分析 + WBClaw 执行协作编排器。将任务发送给 Claude Code Agent 做前置分析设计,再由 WorkBuddy Claw 执行落地,自动生成执行报告。适用于需要深度分析+工程执行的多步骤任务。 It is an AI Agent Skill for Claude Code / OpenClaw, with 49 downloads so far.

How do I install Claw Orchestrator - CC+WBClaw多Agent协作?

Run "/install claw-orchestrator" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Claw Orchestrator - CC+WBClaw多Agent协作 free?

Yes, Claw Orchestrator - CC+WBClaw多Agent协作 is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Claw Orchestrator - CC+WBClaw多Agent协作 support?

Claw Orchestrator - CC+WBClaw多Agent协作 is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Claw Orchestrator - CC+WBClaw多Agent协作?

It is built and maintained by ZenYang-X (@zenyang-x); the current version is v1.0.0.

💬 Comments