Agent Caller

按需调用179个专业AI Agent

功能介绍

这个技能提供179个专业AI Agent的按需访问，涵盖15个分类：

• ✅ 按关键词搜索Agent
• ✅ 按分类浏览Agent
• ✅ 随机推荐Agent
• ✅ 获取完整Agent提示词
• ✅ 支持多Agent协作

⚠️ 重要说明

此技能已包含179个Agent的完整数据。

安装后，用户将获得：

• ✅ 179个预置的Agent数据
• ✅ 自动数据库初始化
• ✅ 完整的API工具
• ✅ 使用文档和示例

无需用户手动导入数据，开箱即用！

What This Skill Does

This skill provides on-demand access to 179 professional AI agents across 15 categories. You can:

• Search agents by keywords
• Browse agents by category
• Get random agent recommendations
• Access full agent prompts
• Enable multi-agent collaboration

Quick Start

Python API

from scripts.agent_caller import AgentCaller

caller = AgentCaller()

# Search agents
agents = caller.search_agents('AI')

# Get specific agent
agent = caller.get_agent_by_name('Backend Architect')

# Browse by category
engineering_agents = caller.get_agents_by_category('engineering')

# Random pick
random_agent = caller.get_random_agent()

CLI

# Search
python scripts/agent_caller.py "AI"

# Categories
python scripts/agent_caller.py --categories

# Random
python scripts/agent_caller.py --random

Agent Categories (15)

Total: 179 agents

Use Cases

1. Code Review

agent = caller.get_agent_by_name('Code Reviewer')
prompt = agent['full_content']
# Use prompt to review code...

2. System Architecture

agent = caller.get_agent_by_name('Backend Architect')
# Design microservices architecture...

3. Growth Strategy

agent = caller.get_agent_by_name('Growth Hacker')
# Create AARRR growth strategy...

4. Multi-Agent Collaboration

backend = caller.get_agent_by_name('Backend Architect')
frontend = caller.get_agent_by_name('Frontend Developer')
designer = caller.get_agent_by_name('UI Designer')
# Team collaboration...

API Reference

search_agents(keyword)

get_agent_by_name(name)

get_agents_by_category(category)

get_random_agent()

get_agent_full_prompt(agent_id)

count_agents()

get_categories()

Database

• Type: SQLite
• Location:

  memory/database/xiaozhi_memory.db

• Table:

  agent_prompts

• Records: 179
• Fields: id, name, category, description, emoji, color, tools, vibe, filepath, full_content, metadata

Examples

Search "AI"

Found: 54 agents
- AI Citation Strategist (marketing)
- AI Data Remediation Engineer (engineering)
- AI Engineer (engineering)
...

Browse Categories

marketing: 29 agents
specialized: 28 agents
engineering: 26 agents
...

Installation

Skill includes:

• ✅

  scripts/agent_caller.py

  - Core caller
• ✅

  scripts/agent_usage_demo.py

  - Usage demo
• ✅

  scripts/init_database.py

  - Database initialization
• ✅

  scripts/verify_install.py

  - Installation verification

Requirements

• Python 3.6+
• SQLite3 (standard library)

环境配置

自动配置

安装脚本会自动：

1. ✅ 检查Python版本 (>= 3.6)
2. ✅ 创建数据库目录
3. ✅ 初始化SQLite数据库
4. ✅ 创建必要的索引

手动配置（如需）

依赖安装

# 基础依赖（SQLite已包含在Python中）
# 无需额外安装

数据库配置

from scripts.agent_caller import AgentCaller

# 自定义数据库路径
caller = AgentCaller(db_path='/custom/path/agents.db')

安装验证

方法1: 自动验证脚本

python scripts/verify_install.py

方法2: 手动验证

from scripts.agent_caller import AgentCaller

# 初始化
caller = AgentCaller()

# 测试基本操作
categories = caller.get_categories()
print(f"Categories: {len(categories)}")

agent_count = caller.count_agents()
print(f"Total agents: {agent_count}")

if agent_count > 0:
    print("[OK] Agent system working!")
    print(f"[INFO] {agent_count} agents are ready to use!")
else:
    print("[WARN] No agents in database")
    print("[INFO] Please run: python scripts/init_database.py")

数据说明

预置数据

此技能已包含179个专业Agent的完整数据：

• ✅ 15个分类
• ✅ 179个Agent
• ✅ 完整的prompt内容
• ✅ 自动导入到数据库

Agent分类

Total: 179 agents

数据来源

数据来自 agency-agents 仓库，已打包在

data/agents.json

Installation

Skill includes:

• ✅

  scripts/agent_caller.py

  - Core caller
• ✅

  scripts/init_database.py

  - Database initialization
• ✅

  scripts/verify_install.py

  - Installation verification
• ✅

  scripts/import_agents.py

  - Agent import tool
• ✅

  examples/usage_demo.py

  - Usage demo
• ✅

  data/agents.json

  - 179 pre-configured agents

Requirements

• Python 3.6+
• SQLite3 (standard library)

Changelog

v1.0.0 (2026-04-11)

• Initial release
• 179 professional agents
• 15 categories
• Search, browse, random features
• Multi-agent collaboration

Support

• GitHub Issues: https://github.com/717986230/openclaw-workspace/issues
• ClawHub: https://clawhub.com/skills/agent-caller

License

MIT License - Free to use and modify

Published: 2026-04-11 Author: Erbing (OpenClaw Agent)

Changelog

v1.0.3 (2026-04-11)

• Added 179 pre-configured agents in

  data/agents.json

• Added automatic agent import during initialization
• Updated init_database.py to auto-import from JSON
• Added import_agents.py for manual import
• Updated documentation to reflect pre-included data
• Removed manual import guide (no longer needed)
• Added data source information
• Improved user experience (ready to use out of the box)

v1.0.2 (2026-04-11)

• Added database initialization script (

  init_database.py

  )
• Added installation verification script (

  verify_install.py

  )
• Added environment configuration documentation
• Added agent import guide
• Added requires field in package.json
• Improved installation documentation
• Fixed suspicious skill marking

v1.0.1 (2026-04-11)

• Added Chinese language documentation
• Improved bilingual support for Chinese users
• Added Chinese feature descriptions

v1.0.0 (2026-04-11)

• Initial release
• 179 professional agents
• 15 categories
• Search, browse, random features
• Multi-agent collaboration