A comprehensive tool for the lifecycle management of AI agents within the OpenClaw ecosystem, from workspace initialization to Telegram routing.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install claw-agent-creator-archit
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install claw-agent-creator-archit using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is OpenClaw Agent Creator?
The OpenClaw Agent Creator provides a structured framework for expanding your Openclaw Skills library by automating the deployment of new agents. This skill handles the intricate details of Linux-based agent hosting, including workspace provisioning, identity formation via Markdown templates, and complex configuration of the OpenClaw gateway. By using this skill, developers can ensure that every new addition to their Openclaw Skills environment follows best practices for stability and performance.
Built on hard-won engineering lessons, it streamlines the setup of Telegram bot routing, model cascades, and self-healing cron jobs. Whether you are building a simple Q&A agent or a complex automated news briefer, this skill ensures that your Openclaw Skills architecture remains modular and collision-free.
OpenClaw Agent Creator Use Cases
- Rapid deployment of new specialized AI agents within the OpenClaw framework.
- Reconfiguring existing Openclaw Skills for better Telegram group integration.
- Establishing automated workflows using self-healing cron job patterns.
- Debugging agent routing and session collisions in multi-agent environments.
How OpenClaw Agent Creator Works
- Requirement Analysis: Define the agent ID, responsibilities, and required model tiers for the new Openclaw Skills entry.
- State Protection: Safely stop the OpenClaw gateway and backup the main configuration file to prevent data corruption.
- Workspace Provisioning: Create dedicated directories for the agent's core files and long-term memory to avoid session collisions.
- Identity Engineering: Write the SOUL.md and IDENTITY.md files using standard templates to define the agent's personality.
- System Integration: Update the openclaw.json configuration with new agent entries, model overrides, and Telegram channel bindings.
- Automation Setup: Inject dynamic cron jobs into jobs.json using specialized prompt patterns that resolve Telegram group IDs at runtime.
- Activation: Restart the gateway and monitor the logs to verify successful registration of the new Openclaw Skills.
OpenClaw Agent Creator Setup
To begin using this skill within your Openclaw Skills environment, follow these steps:
- Stop the gateway process:
openclaw gateway stop
- Create the agent workspace directories:
mkdir -p ~/.openclaw/workspace-<agent_id>/memory
mkdir -p ~/.openclaw/agents/<agent_id>/agent
- Populate the workspace with core Markdown files (SOUL.md, IDENTITY.md) from the templates directory.
- Update the openclaw.json manifest with the agent's metadata and restart the system:
openclaw gateway start
OpenClaw Agent Creator Data Schema & Taxonomy
| Component | Location | Description |
|---|---|---|
| Master Config | ~/.openclaw/openclaw.json |
The central registry for all Openclaw Skills and routing logic. |
| Agent Workspace | ~/.openclaw/workspace-<id>/ |
Contains SOUL.md, IDENTITY.md, and core behavioral definitions. |
| Cron Registry | ~/.openclaw/cron/jobs.json |
Stores scheduled tasks and self-healing prompt engineering. |
| Memory Store | /memory/ |
Persistent JSON and text files used for long-term agent context. |
OpenClaw Agent Creator Advanced Features
- Dynamic Group ID Resolution: Uses shell-integrated preambles to prevent cron job failures during Telegram migrations.
- Model Cascading: Allows granular control over model usage, assigning specific LLMs to different Openclaw Skills.
- Behavioral Modes: Supports complex personality switching for agents operating in different contexts.
- Multi-Agent Routing: Configures unique mention patterns and bindings to ensure messages reach the correct agent.
- Heartbeat Monitoring: Enables or disables periodic agent activity checks based on the specific use case.
SKILL.md
Loading
Related Openclaw Skills
OpenClaw Ecosystem News
A specialized briefing tool that aggregates releases, security advisories, and community trends across the OpenClaw ecosystem.
Arc-shield Secret Sanitizer
Arc-shield is an advanced output filtering skill designed to scan and sanitize AI agent responses, ensuring sensitive credentials and PII never leak to external channels.
Arc Sentinel
A comprehensive security monitoring and infrastructure health toolkit designed for Openclaw Skills agents.
Mixcache Ebook Search Suggester
An AI-driven search assistant that generates optimized Mixcache ebook discovery links based on user topics, authors, or genres.
Raspberry Pi CPU Temperature Monitor
A thermal monitoring and alerting system for Raspberry Pi that integrates hardware sensors with Openclaw Skills for real-time notifications.
S.I.S. - Sovereign Intelligence System
S.I.S. provides equilibrium-native reasoning for OpenClaw assistants, ensuring every operation maintains logical balance and self-validating coherence.
Featured*
