Agent Maker

v1.0.0

辅助创建 OpenClaw Agent 的技能。通过对话引导用户描述需求,自动生成新的 Agent 配置文件。

0· 162·0 current·0 all-time

Install

OpenClaw Prompt Flow

Install with OpenClaw

Best for remote or guided setup. Copy the exact prompt, then paste it into OpenClaw for squallfire/agent-maker.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "Agent Maker" (squallfire/agent-maker) from ClawHub.
Skill page: https://clawhub.ai/squallfire/agent-maker
Keep the work scoped to this skill only.
After install, inspect the skill metadata and help me finish setup.
Use only the metadata you can verify from ClawHub; do not invent missing requirements.
Ask before making any broader environment changes.

Command Line

CLI Commands

Use the direct CLI path if you want to install manually and keep every step visible.

OpenClaw CLI

Bare skill slug

openclaw skills install agent-maker

ClawHub CLI

Package manager switcher

npx clawhub@latest install agent-maker
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
Purpose & Capability
The name/description match the included tooling: the scripts create agent config files and workspaces. Minor inconsistencies in documentation/path names exist (SKILL.md mentions ~/.openclaw/workspace/agents/... whereas the scripts use ~/.openclaw/agents/ and ~/.openclaw/workspaces/). README.md also states "creating Agent requires administrator permissions" though the scripts operate on user HOME paths and do not require sudo. These are documentation mismatches, not functional red flags.
Instruction Scope
SKILL.md instructs runtime execution of ./tools/create-agent.sh with user-supplied arguments. The scripts only create directories and text files under the user's HOME, and list/validate agent directories; they do not perform network calls, read unrelated system credentials, or transmit data externally.
Install Mechanism
No install spec or remote downloads; this is an instruction-only skill with bundled local shell scripts. Nothing is fetched from external URLs or written outside the user's home directory by the provided scripts.
Credentials
The skill declares no required environment variables or credentials and the scripts only rely on standard environment variables like HOME and common shell utilities. No secret-like environment variables are requested.
Persistence & Privilege
The skill creates persistent files under ~/.openclaw/agents and workspaces under the user's HOME (expected for its purpose). always:false is set. Ensure you are comfortable with these persistent files; the skill does not modify other skills' configurations or request elevated privileges.
Assessment
This skill appears to do what it claims: locally generate agent config folders and files. Before installing/running it, review the create-agent.sh contents (they are included) and note the following: (1) the documentation contains inconsistent path names—expect files under ~/.openclaw/agents and ~/.openclaw/workspaces as the script implements; (2) README mentions admin permissions but the script does not use sudo—no elevated privileges are required unless you choose system locations; (3) the tool will create persistent files in your home directory—verify the generated files and workspace locations after creation; and (4) the skill references linking to other Skills (soft links/references) — confirm any linked Skill names you provide are trusted. If you want extra caution, run the script in a disposable/test account or inspect the created files before using the new Agent in production.

Like a lobster shell, security has layers — review code before you run it.

latestvk976qwvzy231599as6qxj2bj4n8333k7
162downloads
0stars
1versions
Updated 1mo ago
v1.0.0
MIT-0

🎯 使用说明

当你需要创建一个新的 OpenClaw Agent 时,可以直接 @agent-maker 并描述你的需求。我会引导你完成 Agent 的配置,并生成完整的 Agent 定义文件。

触发方式

  • @agent-maker 我想创建一个运维助手
  • @agent-maker 帮我生成一个客服机器人
  • @agent-maker 创建数据分析师 Agent

📋 Workflow

1. 需求收集

通过多轮对话收集以下信息:

字段说明示例
Agent 名称英文小写+短横线ops-agent
角色描述一句话概括资深运维工程师
核心职责3-5 项主要工作安装部署、安全巡检、木马查杀
关联 Skill需要调用的现有技能deploy-service, security-audit
运行规则触发条件/定时任务每天凌晨2点自动巡检
输出格式结果返回方式Markdown 报告

2. 生成配置文件

根据收集的信息,在 OpenClaw 的 agents 目录下创建: ~/.openclaw/workspace/agents/{agent_name}/ ├── SOUL.md # Agent 核心定义 ├── README.md # 使用说明 └── skills/ # 关联 Skill 的软链接

3. 注册与反馈

  • 生成 Agent 配置文件
  • 提示用户重启网关使生效
  • 返回使用方式和示例

📝 对话模板

场景一:创建运维 Agent

用户:@agent-maker 我想创建一个运维助手 Agent Maker:好的,我们来一步步配置这个运维 Agent。

请为这个 Agent 起一个英文名称(小写+短横线): 用户:ops-agent

用一句话描述它的角色: 用户:资深运维工程师,负责服务器部署和安全维护

它的核心职责有哪些?(每行一项)

用户:每天扫描系统漏洞 用户:发现木马立即隔离 用户:监控系统资源

需要调用哪些现有 Skill?(如没有则留空) 用户:deploy-service, security-audit, malware-kill

是否有特殊运行规则?(如定时任务) 用户:每天凌晨2点自动执行安全巡检

✅ Agent "ops-agent" 创建成功!

配置文件:~/.openclaw/workspace/agents/ops-agent/SOUL.md

关联 Skill:deploy-service, security-audit, malware-kill

使用方式:@ops-agent [你的指令]

请执行 openclaw gateway restart 使新 Agent 生效。

场景二:创建客服 Agent

用户:@agent-maker 帮我生成一个客服机器人 Agent Maker:好的,我们来配置客服 Agent。

Agent 名称(英文小写+短横线): 用户:customer-service

角色描述: 用户:友好耐心的客服助手,解答用户常见问题

核心职责: 用户:回答产品咨询 用户:处理售后问题 用户:引导用户操作

关联 Skill: 用户:search-docs, ticket-system

运行规则: 用户:7x24小时自动响应,复杂问题转人工

✅ Agent "customer-service" 创建成功!

🛠️ 高级功能

支持模板创建

如果用户想快速创建常见类型的 Agent,可以直接选择模板:

  • @agent-maker 用运维模板创建一个 Agent
  • @agent-maker 用客服模板

支持 Skill 推荐

根据用户描述的职责,自动推荐相关的 Skill: 检测到您需要“安全巡检”功能,推荐安装以下 Skill:

security-audit:基础安全扫描

malware-kill:木马查杀

log-monitor:日志监控

支持配置文件预览

生成前可以让用户预览 SOUL.md 内容,确认无误后再保存。

📌 输出格式规范

生成的 SOUL.md 文件格式:


name: {agent_name} description: {role_description} created_at: {timestamp}

职责

{bullet_points_of_duties}

可用 Skill

{list_of_skills}

运行规则

{run_rules}

使用示例

@{agent_name} [你的指令]

Comments

Loading comments...