ClawSkillShield 🛡️\r

\r Local-first security scanner for OpenClaw/ClawHub skills.\r \r

What It Does\r

\r

• Static analysis for security risks and malware patterns\r
• Detects:\r
  • Hardcoded secrets (API keys, credentials, private keys)\r
  • Risky imports (os, subprocess, socket, ctypes)\r
  • Dangerous calls (eval(), exec(), open())\r
  • Obfuscation (base64 blobs, suspicious encoding)\r
  • Hardcoded IPs\r
• Risk scoring (0–10) + detailed threat reports\r
• Quarantine high-risk skills automatically\r \r

Dual-Use Design\r

\r

• CLI for humans: Quick safety checks before installing skills\r
• Agent API: Importable functions for autonomous agents/Moltbots to proactively scan and quarantine risky skills (essential post-ClawHavoc)\r \r

Quick Start\r

\r

CLI (Humans)\r

pip install -e .\r
clawskillshield scan-local /path/to/skill\r
clawskillshield quarantine /path/to/skill\r
```\r
\r
### Python API (Agents)\r
```python\r
from clawskillshield import scan_local, quarantine\r
\r
threats = scan_local("/path/to/skill")\r
if risk_score \x3C 4:  # HIGH RISK\r
    quarantine("/path/to/skill")\r
```\r
\r
## Zero Dependencies\r
Pure Python. No network calls. Runs entirely locally.\r
\r
## Why This Matters\r
ClawHavoc demonstrated how easily malicious skills can slip into the ecosystem. ClawSkillShield provides a trusted, open-source defense layer—audit the code, run offline, stay safe.\r
\r
---\r
\r
**GitHub**: https://github.com/AbYousef739/clawskillshield  \r
**License**: MIT  \r
**Author**: Ab Yousef  \r
**Contact**: [email protected]\r