ClawShell is a security-focused tool that intercepts shell commands and requires push notification approval for high-risk operations.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install clawshell-0-1-0
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 clawshell-0-1-0 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 ClawShell?
ClawShell serves as a critical human-in-the-loop security layer designed specifically for Openclaw Skills. It functions by intercepting shell commands before they are executed, performing a real-time risk analysis, and mandating explicit user approval for any operation deemed dangerous. This ensures that even the most powerful AI agents remain under human control, preventing accidental data loss or unauthorized system changes.
By integrating this skill, developers can leverage the full power of Openclaw Skills while maintaining a robust security posture. It replaces the standard bash interface with a secure wrapper, allowing for a defense-in-depth approach that complements existing sandboxing techniques and provides peace of mind when deploying autonomous agents.
ClawShell Use Cases
- Preventing destructive commands such as rm -rf from running without explicit verification.
- Intercepting and approving external network requests or credential access attempts.
- Auditing all shell activity within an AI-driven development environment.
- Managing high-stakes operations like git pushes or package installations through remote approval.
How ClawShell Works
- The AI agent calls the clawshell_bash tool instead of the standard system bash.
- ClawShell analyzes the command against a comprehensive set of built-in and configurable risk rules.
- The command is categorized into risk levels: Critical (blocked), High (requires approval), Medium (logged/allowed), or Low (allowed).
- For High-risk commands, a push notification is sent to the user via Pushover or Telegram.
- The execution blocks until the user provides approval or the configured timeout is reached.
- All decisions and command outputs are recorded in a centralized JSONL log for security auditing within the Openclaw Skills ecosystem.
ClawShell Setup
To begin using ClawShell with your Openclaw Skills, navigate to the skill directory and install the necessary dependencies:
cd /app/workspace/skills/clawshell
npm install
Configure your notification provider by adding your API keys to the .env file:
CLAWSHELL_PUSHOVER_USER=your-user-key
CLAWSHELL_PUSHOVER_TOKEN=your-app-token
Finally, update your TOOLS.md to instruct the agent to utilize clawshell_bash for all shell execution tasks.
ClawShell Data Schema & Taxonomy
ClawShell organizes its security data and configuration through the following schema:
| Component | Type | Description |
|---|---|---|
| logs/clawshell.jsonl | Audit Log | A line-delimited JSON file containing command history, risk levels, and approval status. |
| config.yaml | Configuration | Defines custom blocklists and allowlists using regex or glob patterns. |
| CLAWSHELL_BLOCKLIST | Env Var | A comma-separated list of additional commands to be automatically rejected. |
| CLAWSHELL_ALLOWLIST | Env Var | A comma-separated list of trusted commands that bypass high-risk checks. |
ClawShell Advanced Features
- Multi-channel notification support including Pushover and Telegram for flexible remote approvals.
- Configurable timeout settings to automatically reject pending requests after a specified duration.
- Custom rule engine allowing for granular control over command patterns using regular expressions.
- Seamless integration with Openclaw Skills audit logs for comprehensive security monitoring and forensic analysis.
SKILL.md
Loading
Related Openclaw Skills
Filesystem Management
A comprehensive filesystem toolkit for AI agents that enables advanced file listing, pattern-based searching, and safe batch operations.
Signal Integration
A full-featured Signal messenger bridge for AI agents featuring encrypted text, voice transcription, and role-based security.
Codex CN Bridge
A protocol conversion bridge that allows the OpenAI Codex CLI to utilize high-performance domestic Chinese AI models.
Weather Checker
A professional Python-based command-line utility providing global weather forecasts, precipitation data, and temperature tracking with emoji-rich formatting.
SafeExec
Protect your system with an automated command interception and human-in-the-loop approval workflow for AI agent shell operations.
Web Search Skill
A high-performance skill that enables AI agents to query the internet for real-time information using the DuckDuckGo API.
Featured*
