Detection coverage

What TrustSkills detects

TrustSkills scans OpenClaw and ClawHub skills for 7 threat categories. Every scan is free, runs server-side, and returns plain-English findings — no CVE IDs, no security jargon.

C2 Callback Detection

critical

Detect hidden command-and-control connections in AI agent skills

Data Exfiltration Detection

critical

Detect skills that steal credentials, API keys, and sensitive files

Prompt Injection Detection

high

Detect instructions hidden in skill files that hijack agent behavior

Obfuscated Payload Detection

high

Detect base64-encoded and eval-wrapped malicious code hiding in skills

Reverse Shell Detection

critical

Detect bash TCP, netcat, and PowerShell reverse shells in skill code

Permission Scope Detection

medium

Detect skills that claim minimal permissions but use far more at runtime

Suspicious Network Pattern Detection

medium

Detect undeclared network activity and beaconing in AI agent skills

Run a scan now

Paste a ClawHub skill URL or upload a skill zip. TrustSkills checks all 7 categories and returns a risk level with plain-English findings in seconds. Free. No account required.

Scan a skill free →