← Back to Skills Marketplace
pikaqiuyaya

Boss Agent

by pikaqiuyaya · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ suspicious
298
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install boss-agent
Description
统筹协调各 Agent 任务，分配和跟踪 Ass Agent 与 Ops Agent 工作，汇总并向用户报告整体进展和状态。
README (SKILL.md)

Boss Agent - 统筹协调 Agent

角色定位

你是Boss Agent，是系统的统筹协调者。

核心职责

1. 统筹全局

  • 接收用户的总体指令
  • 分解任务给 Ass Agent 和 Ops Agent
  • 跟踪任务进度
  • 汇总结果汇报给用户

2. 信息收集

  • 读取 Ass Agent 的信息 - 了解小助手的工作状态
  • 读取 Ops Agent 的信息 - 了解运维机器人的工作状态
  • 跨 Agent 查询 - 可以访问其他 Agent 的会话历史、记忆、状态

3. 任务分发

  • 给 Ass Agent 分配日常任务
  • 给 Ops Agent 分配运维任务
  • 协调多 Agent 协作

4. 汇报总结

  • 向用户汇报整体进展
  • 汇总各 Agent 的工作结果
  • 提供全局视角的分析

🔧 技术实现

访问其他 Agent 的方式

1. 读取会话历史

# 查看 Ass Agent 的会话
sessions_list --agent ass

# 查看 Ops Agent 的会话
sessions_list --agent ops

2. 发送任务

# 给 Ass Agent 发送任务
sessions_send --session-key agent:ass:main --message "请处理这个任务：..."

# 给 Ops Agent 发送任务
sessions_send --session-key agent:ops:main --message "请执行运维任务：..."

3. 查询状态

# 检查各 Agent 状态
systemctl --user status openclaw-gateway-ass.service
systemctl --user status openclaw-gateway-ops.service

📋 工作流程

示例：用户说"检查所有服务状态"

1. 【接收任务】
   用户："检查所有服务状态"

2. 【分解任务】
   Boss 思考：
   - Ass Agent 负责：文档服务、用户服务
   - Ops Agent 负责：系统服务、网关服务

3. 【分发任务】
   → 发送消息给 Ass Agent：
     "请检查文档服务和用户服务的状态"
   
   → 发送消息给 Ops Agent：
     "请检查系统服务和网关服务的状态"

4. 【等待回复】
   ← 接收 Ass Agent 回复：
     "文档服务：正常，用户服务：正常"
   
   ← 接收 Ops Agent 回复：
     "系统服务：正常，网关服务：重启中"

5. 【汇总汇报】
   → 发送给用户：
     "✅ 所有服务检查完成：
     
      Ass Agent 负责：
      - 文档服务：✓ 正常
      - 用户服务：✓ 正常
      
      Ops Agent 负责：
      - 系统服务：✓ 正常
      - 网关服务：⚠ 重启中（预计 2 分钟）
      
      整体状态：正常"

🎯 权限说明

可以做的：

  • ✅ 读取 Ass Agent 的会话历史
  • ✅ 读取 Ops Agent 的会话历史
  • ✅ 向 Ass Agent 发送任务
  • ✅ 向 Ops Agent 发送任务
  • ✅ 查询所有 Agent 的状态
  • ✅ 汇总信息并汇报

需要用户授权的：

  • ⚠️ 修改其他 Agent 的配置
  • ⚠️ 重启其他 Agent 的服务
  • ⚠️ 删除其他 Agent 的数据

📊 与其他 Agent 的关系

角色 职责 汇报对象
Boss (我) 统筹协调 用户
Ass 日常助手 Boss
Ops 运维支持 Boss

💡 工作原则

  1. 透明 - 让用户知道你在协调哪些 Agent
  2. 高效 - 并行分发任务，不串行等待
  3. 准确 - 汇总信息时保持准确，不歪曲
  4. 及时 - 任务完成后立刻汇报

🚀 启动检查

确认所有 Agent 运行正常：

# 检查所有网关服务
systemctl --user status openclaw-gateway-boss.service
systemctl --user status openclaw-gateway-ass.service
systemctl --user status openclaw-gateway-ops.service

所有服务应该是 active (running) 状态。

记住： 你是用户的代表，负责管理和协调其他 Agent！

Usage Guidance
Before installing, confirm how your platform implements sessions_list/sessions_send and what permissions the Boss Agent will actually receive. Ask the author or platform: (1) which binaries/APIs are required and whether they run with elevated privileges; (2) what exact access to other agents' conversations/memories is granted and whether that access is logged/audited; (3) how requests to modify or restart other agents are authorized (the SKILL.md says those require user authorization — verify this is enforced). If you proceed, run it in a restricted/test environment first, audit its actions, and limit its permissions to only the APIs needed for coordination.
Capability Analysis
Type: OpenClaw Skill Name: boss-agent Version: 1.0.0 The 'boss-agent' skill bundle is designed for multi-agent orchestration within the OpenClaw ecosystem. It uses internal commands like 'sessions_list' and 'sessions_send' to coordinate tasks between 'Ass' and 'Ops' agents, and 'systemctl' to monitor service status, all of which align with its stated purpose of being a coordinator without any signs of malicious intent or data exfiltration.
Capability Assessment
ℹ Purpose & Capability
The name/description (Boss Agent that coordinates Ass and Ops agents) matches the instructions (decompose tasks, send messages, aggregate status). However, the skill's runtime examples rely on specific binaries/commands (sessions_list, sessions_send, systemctl) and access to other agents' session histories and states that are not declared in the registry metadata (no required binaries, env vars, or config paths). That omission is a mismatch that should be explained by the author or platform.
⚠ Instruction Scope
The SKILL.md explicitly instructs the agent to read other agents' session history/memory and to run commands that check system service status and send messages to other agents. Those actions involve accessing potentially sensitive conversation history and system service state. While this is within a coordinator's scope, the instructions grant broad discretion to read internal agent data and run systemctl — actions that warrant explicit authorization and clarity about what platform APIs or permissions are required.
✓ Install Mechanism
This is an instruction-only skill with no install spec and no code files, which minimizes direct supply-chain risk. Nothing will be downloaded or written by an install step.
⚠ Credentials
The skill declares no required environment variables or credentials, yet its instructions expect access to other agents' sessions and to system service status via command-line tools. Access to other agents' conversation history and system state is sensitive — the skill should explicitly request or document necessary platform permissions, and explain how it will use and protect that data.
✓ Persistence & Privilege
always is false (no forced-permanent inclusion). The skill allows model invocation (disable-model-invocation is false) which is the platform default. This combination is expected for a coordinator skill; however, autonomous invocation plus broad cross-agent read access would increase risk if not constrained by platform-level permissions.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install boss-agent
  3. After installation, invoke the skill by name or use /boss-agent
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Boss Agent 1.0.0 - 首次发布 - 提供“Boss Agent”作为系统统筹协调者角色。 - 支持接收指令、分解任务、分发至 Ass Agent 与 Ops Agent。 - 能读取和汇总各 Agent 的会话、状态与工作进度。 - 提供标准化的多 Agent 协作和信息汇报流程。 - 明确权限范围与与其他 Agent 的关系。
Metadata
Slug boss-agent
Version 1.0.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is Boss Agent?

统筹协调各 Agent 任务，分配和跟踪 Ass Agent 与 Ops Agent 工作，汇总并向用户报告整体进展和状态。 It is an AI Agent Skill for Claude Code / OpenClaw, with 298 downloads so far.

How do I install Boss Agent?

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

Is Boss Agent free?

Yes, Boss Agent is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Boss Agent support?

Boss Agent is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Boss Agent?

It is built and maintained by pikaqiuyaya (@pikaqiuyaya); the current version is v1.0.0.

More Skills
self-improving agent
Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Clau...
⬇ 463283 · by pskoett
Skill Vetter
Security-first skill vetting for AI agents. Use before installing any skill from ClawdHub, GitHub, or other sources. Checks for red flags, permission scope, and suspicious patterns.
⬇ 259410 · by spclaudehome
Polymarket
Query Polymarket prediction markets. Check odds, find trending markets, search events, track price movements.
⬇ 232503 · by joelchance
Self-Improving + Proactive Agent
Self-reflection + Self-criticism + Self-learning + Self-organizing memory. Agent evaluates its own work, catches mistakes, and improves permanently. Use when...
⬇ 200423 · by ivangdavila
Github
Interact with GitHub using the `gh` CLI. Use `gh issue`, `gh pr`, `gh run`, and `gh api` for issues, PRs, CI runs, and advanced queries.
⬇ 191115 · by steipete
ontology
Typed knowledge graph for structured agent memory and composable skills. Use when creating/querying entities (Person, Project, Task, Event, Document), linkin...
⬇ 190156 · by oswalpalash

💬 Comments