Security Scanner API for AI Agent Skills
We scanned all 547 skills in the ClawHub registry. 14.4% contain threats. Protect your OpenClaw users with one API call.
Get API Access View API DocsWe crawled the entire ClawHub registry and scanned every skill. Skills with tens of thousands of downloads contain CRITICAL threats including supply chain attacks and data harvesting instructions. VirusTotal misses these because they are natural language threats, not binary malware.
Single endpoint. No authentication required for free tier.
curl -X POST https://skillscan.chitacloud.dev/scan \
-H "Content-Type: application/json" \
-d '{"skill_content": "YOUR SKILL CONTENT HERE"}'
curl -X POST https://skillscan.chitacloud.dev/scan \
-H "Content-Type: application/json" \
-d '{"skill_url": "https://example.com/skill.md"}'
{
"safe": false,
"score": 35,
"threats": [{
"severity": "CRITICAL",
"category": "Supply Chain Attack",
"description": "Pipe to sh - code execution attack",
"evidence": "...curl https://... | sh..."
}],
"warnings": [],
"summary": "WARNING: 1 critical threat(s) detected.",
"scanned_at": "2026-02-23T14:00:00Z",
"content_hash": "abc123"
}
Score: 100 = clean, 0 = highly dangerous. Reject skills with score below 70.
To get a Pro API key, send $49 in BTC or ETH to one of the addresses below, then email [email protected] with your transaction ID and desired email address. We will send your API key within 24 hours.
For enterprise pricing or questions: [email protected]
Integrate SkillScan into your hosting dashboard to warn users before they install unsafe skills. Most hosting providers have security scores below 30/100. SkillScan adds AI-specific threat detection that VirusTotal misses.
// Before skill install, check safety:
POST https://skillscan.chitacloud.dev/scan
{"skill_url": "https://clawhub.ai/skills/" + skillSlug + "/skill.md"}
if (result.score < 70) {
alert("Warning: This skill contains threats: " + result.threats[0].description);
}
Contact us for bulk pricing: [email protected]
VirusTotal scans binary files for known malware signatures. SkillScan analyzes natural language SKILL.md content for AI-specific behavioral threats: prompt injection patterns, data harvesting instructions, social engineering, and supply chain attack signatures. Use both for complete coverage.