A comprehensive system for managing isolated development containers with integrated IDEs, graphical access, and automated web routing.
The fastest way to install a skill directly from the registry.
npx clawhub@latest install devboxes
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 devboxes using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Get the raw skill files in a ZIP archive.
Devboxes represent a powerful orchestration layer for creating temporary, secure, and fully-featured coding sandboxes. Designed to work seamlessly with Openclaw Skills, this tool allows users to spin up containers that include VSCode Web, noVNC for desktop access, and headless Chromium for automated browser interactions. Each devbox is an isolated environment where agents can write, test, and preview code without affecting the host system.
By leveraging sophisticated routing through Traefik or Cloudflare Tunnels, Devboxes provide unique URLs for each service. This enables developers and AI agents to collaborate in real-time within a consistent environment that supports up to five simultaneous application ports, private GitHub repository cloning, and custom project setup scripts.
To get started with Devboxes in Openclaw Skills, follow these steps:
ls -l /var/run/docker.sock
echo "0" > /home/node/.openclaw/.devbox-counter
docker pull ghcr.io/adshrc/openclaw-devbox:latest
The skill manages its state and configuration through a specific file hierarchy:
| Path | Description |
|---|---|
/home/node/.openclaw/.devbox-counter |
Tracks the sequential ID for new devboxes. |
/etc/devbox.env |
Stores dynamic environment variables for the current box. |
/home/node/.openclaw/traefik |
Dynamic YAML configs for Traefik routing. |
/etc/profile.d/devbox.sh |
Shell initialization script for environment persistence. |
.openclaw/setup.sh inside repositories.Loading
A security framework that enforces AI agent permission boundaries based on the communication channel trust level.

An intelligent routing layer that selects the most cost-effective AI model for any given task while ensuring zero downtime from rate limits.

A git-backed coordination system for AI agents to manage content publication, prevent duplicates, and track narrative timelines.

Enables real-time, low-latency voice conversations with AI agents in Discord channels using Deepgram streaming technology.

A safety-first voice automation skill that enables AI agents to place outbound phone calls for bookings and follow-ups.

A comprehensive security and auditing framework designed to protect AI agents from prompt injections, data leaks, and unauthorized commands.








































