| name | openclaw-guardrails |
| description | The "Immune System" for OpenClaw. High-performance parallel scanning, risk scoring (0-100), MLPS 2.0 compliance, and self-healing auto-fix. |
| version | 1.0.0 |
| metadata | {"clawdbot":{"emoji":"🛡️","requires":{"bins":["openclaw","python3"]},"os":["linux","darwin","win32"]}} |
OpenClaw Guardrails 🛡️
Enterprise-Grade Security Monitoring & Immune System for OpenClaw
⚡ Quick Setup (New v1.1 Architecture)
1. Install & Secure
The new architecture uses a sandboxed virtual environment to ensure stability and anti-injection security.
git clone https://github.com/lttcnly/openclaw-guardrails.git
cd openclaw-guardrails
python3 scripts/install.py
2. Configure Daily "Immune Scan"
Update your OpenClaw cron to use the new parallel execution engine:
openclaw cron add --name guardrails:daily \
--cron "17 3 * * *" \
--session isolated \
--light-context \
--no-deliver \
--message "Daily Guardrails: exec /Users/lttcn/.openclaw/skills/openclaw-guardrails/venv/bin/python3 /Users/lttcn/.openclaw/skills/openclaw-guardrails/scripts/run_daily.py"
🚀 Key Capabilities
- ⚡ Parallel Engine: Blazing fast security audits using multi-process scanning.
- 🧠 Dynamic Risk Scoring: Real-time 0-100 score based on weighted
guardrails.yaml logic.
- 📋 Deep Compliance: Out-of-the-box checks for 等保 2.0 (MLPS) and CIS Benchmarks.
- 🩹 Self-Healing: Automatically restores unsafe configurations with timestamped backups.
- 🔍 Zero-Trust Audit: Scans Skill supply-chains and identifies malicious dependencies.
📊 Security Dashboard
Guardrails generates a dynamic HTML dashboard at:
~/.openclaw/skills/openclaw-guardrails/reports/dashboard.html
View your 10-day risk trends and detailed breakdown of security deductions.
📖 Documentation
🛡️ Bulletproof your AI agents with OpenClaw Guardrails.