A powerful management tool for spawning, monitoring, and steering specialized sub-intelligent entities to handle complex workflows.
The fastest way to install a skill directly from the registry.
npx clawhub@latest install create-subagent
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
Copy this prompt to OpenClaw to install it automatically.
Help me install create-subagent using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Get the raw skill files in a ZIP archive.
The SubAgent Creator is a specialized tool designed to extend the capabilities of your primary AI agent by delegating complex tasks to dedicated sub-entities. By utilizing this component of Openclaw Skills, developers can spawn multiple agents with specific system prompts, toolsets, and lifecycle behaviors. This allows for a modular approach to problem-solving where specialized agents handle development, research, writing, or data analysis in parallel.
This skill provides the administrative layer needed to maintain a multi-agent environment. It handles everything from the initial spawning process using templates to real-time steering and logging. By integrating these Openclaw Skills into your workflow, you can move beyond single-threaded task execution and build a collaborative team of AI agents that work autonomously toward a common goal.
To enable sub-agent spawning, you must configure your Openclaw Gateway settings. Update your openclaw-config.json to define permitted agents and concurrency limits.
{
"agents": {
"list": [
{
"id": "main",
"subagents": { "allowAgents": ["main", "developer"] }
}
],
"defaults": {
"subagents": {
"maxConcurrent": 8,
"maxSpawnDepth": 2
}
}
}
}
After updating the configuration, restart your gateway to apply the changes for your Openclaw Skills setup.
The skill organizes sub-agents using a structured metadata taxonomy to ensure trackability and resource management.
| Property | Type | Description |
|---|---|---|
| label | String | A concise identifier for the agent instance (e.g., dev-agent) |
| task | String | The system instructions and role definition for the SubAgent |
| mode | String | Execution strategy: 'run' for tasks or 'session' for threads |
| cleanup | String | Retention policy for the agent session (default: 'keep') |
| run_id | Integer | The unique numerical identifier assigned during spawning |
Loading
A security-focused tool that audits Openclaw Skills for vulnerabilities, suspicious API calls, and malicious patterns prior to installation.

A standardized, high-integrity PR review and merge process designed for task-driven AI agent collaboration.

A structured, quality-first development lifecycle for AI agents involving rigorous planning, test-driven development, and pull request reviews.

Empowers AI agents to send authentic, playable voice messages directly to Feishu users instead of standard file attachments.

A powerful orchestration skill to spawn and manage specialized subordinate agents for parallel task execution.

A comprehensive scholarly research integration for searching databases, downloading PDFs, and automating citation management.








































