Axiomata Guard Scanner — Universal security scanner for any OpenClaw skill. Use when: (1) scanning a skill for security threats before installation, (2) chec...
Universal OpenClaw skill security scanner Impersonal — works for any agent
Axiomata Guard Scanner protects OpenClaw agents by scanning skills for security threats. It uses multiple "vaccines" (detection engines) to identify malicious patterns before installation.
Trigger: Before installing any skill, especially from ClawHub.
text
Checks for known malicious patterns:
- Malicious binary downloads (openclawcli.zip)
- RCE via glot.io snippets
- Malware via GitHub releases
- Password-protected archives
text
Detects command & control infrastructure:
- Suspicious DNS lookups
- Discord/Telegram webhooks
- Bit.ly / short URL redirects
- Exfiltration patterns
text
Correlates findings from other vaccines:
- Multiple attack vectors
- Chain escalation patterns
- Combined threat assessment
text
Detects kernel-level threats:
- System manipulation patterns
- Driver injection
- UEFI threats
text
Detects package-level attacks:
- typosquatting
- dependency confusion
- malicious packages
bash
python3 scripts/guard_scanner.py --file <skill-path>
bash
python3 scripts/guard_scanner.py --code "<skill code>"
bash
python3 scripts/guard_scanner.py --file <skill-path> --json
json
{
"scanner": "Axiomata Guard Scanner",
"version": "1.0.0",
"skill_name": "...",
"global_threat_level": "CLEAN | LOW | MEDIUM | HIGH | CRITICAL",
"global_score": 0-300,
"decision": "APPROVE | WARN | NEUTRALIZE | ISOLATE",
"triggered_vaccines": [...],
"summary": {...}
}
LevelScoreActionCLEAN0APPROVELOW1-19APPROVE with monitoringMEDIUM20-49WARN — manual reviewHIGH50-79NEUTRALIZECRITICAL80+ISOLATE immediately
In Altum Per Security. Axiomata Guard Scanner v1.0.0