安全扫描（ClawHub）

• VirusTotal ：可疑
• OpenClaw ：可疑

OpenClaw 评估

The package claims to centrally manage and route agents but is a prototype that mostly manipulates local markdown and mock data; there are inconsistencies (node is required by SKILL.md but metadata says no binaries, the code doesn't actually call agent platform APIs yet) that make its stated capabilities unclear.

安装（复制给龙虾 AI）

将下方整段复制到龙虾中文库对话中，由龙虾按 SKILL.md 完成安装。

请把本段交给龙虾中文库（龙虾 AI）执行：为本机安装 OpenClaw 技能「Agent manager」。简介：Manage Clawdbot agents: discover, profile, track capabilities, define routing h…。
请 fetch 以下地址读取 SKILL.md 并按文档完成安装：https://raw.githubusercontent.com/openclaw/skills/refs/heads/main/skills/agentandbot-design/agents-manager/SKILL.md
（来源：yingzhi8.cn 技能库）

SKILL.md

打开原始 SKILL.md（GitHub raw）

---
name: agents-manager
description: Manage Clawdbot agents: discover, profile, track capabilities, define routing hierarchy, and assign tasks.
homepage: https://www.clawhub.com/skills/agents-manager
metadata: {"openclaw":{"emoji":"🕵️","requires":{"bins":["node"]}}}
---

# Agents Manager

Manage all Clawdbot agents centrally: discover available agents, profile their capabilities, define routing hierarchy (who can assign to whom, who reports to whom), and intelligently route new tasks through the escalation chain.

## 🤖 For Agents

### System Prompt (Copy-Paste)
Give this prompt to any agent to teach them this skill:
> "You have access to the `agents-manager` skill. Use it to discover peers (`scan_agents`), check permissions (`can_assign.js`), and route tasks. Always follow the Handshake Protocol: Check `requires_approval` before assigning. If true, ask me first."

### Core Rules
1. **Discovery:** Always check `scan_agents.js` before assuming an agent exists.
2. **Permission:** Never assign a task without checking `can_assign.js` or `agent-registry.md`.
3. **protocol:**
   - IF `requires_approval` is FALSE -> Assign directly.
   - IF `requires_approval` is TRUE -> Ask supervisor (Human or Agent).

## 👤 For Humans

### Quick Start
| Goal | Command |
|------|---------|
| **Setup** | `node scripts/setup_wizard.js` (Run this first!) |
| **List** | `node scripts/scan_agents.js` |
| **Health** | `node scripts/health_check.js` |
| **Stats** | `node scripts/log_analyzer.js` |

### 1. Agent Discovery & Profiling
List and profile all agents to understand their capabilities and routing configuration.

```bash
# List all agents
node {baseDir}/scripts/scan_agents.js

# Profile specific agent
node {baseDir}/scripts/generate_card.js <agent_id>
```

### 2. Validation & Health
Ensure your agent ecosystem is healthy and valid.

```bash
# Validate registry integrity
node {baseDir}/scripts/validate_registry.js

# Check permissions (Agent A -> Agent B)
node {baseDir}/scripts/can_assign.js <source_id> <target_id>

# Visualize hierarchy
node {baseDir}/scripts/visualize_agents.js
```

### 3. Task Routing & Escalation
Define how tasks flow between agents using `references/task-routing-rules.md`.

- **Direct:** Agent → Agent (if `can_assign_to` allows)
- **Handshake:** Request approval if `requires_approval` is true.
- **Escalation:** Helper → Supervisor → Human

## Resources

- **[agent-profile-schema.md](references/agent-profile-schema.md)**: Standard profile with routing & card fields.
- **[agent-registry.md](references/agent-registry.md)**: Live registry of all agents.
- **[task-routing-rules.md](references/task-routing-rules.md)**: Decision matrix and handshake protocol.

## Scripts

- `scan_agents.js`: Discovery tool
- `validate_registry.js`: Schema validator
- `can_assign.js`: Permission checker
- `generate_card.js`: Agent card generator
- `visualize_agents.js`: Hierarchy visualizer
- `scan_agents.js`: Discovery tool
- `validate_registry.js`: Schema validator
- `can_assign.js`: Permission checker
- `generate_card.js`: Agent card generator
- `visualize_agents.js`: Hierarchy visualizer
- `health_check.js`: Status monitor (Healthy/Slow/Offline)
- `log_analyzer.js`: Performance stats (Jobs/Success Rate)
- `setup_wizard.js`: Interactive configuration tool