Skip to content
System Telemetry

Scanning Statistics

Real-time transparency into the state of AI agent skill security.

Looking for the snapshot research report? Read the Intelligence Report →

10,242

Skills Scanned (Live)

96

Avg Trust Score

45,300

Total Findings

9,344

Certified

Badge Distribution

CERTIFIED
9344 (91%)
CONDITIONAL
495 (5%)
SUSPICIOUS
134 (1%)
REJECTED
269 (3%)

Scan Sources

ClawHub Registry
6,595 (64%)
Public API● api
2,600 (25%)
Website Submit Page● website
1,044 (10%)
Admin
3 (0%)

1,044 skills submitted from the website

2,600 skills submitted via API clients

Most Common Findings

lowUnknown external reference24245×
lowNo explicit safety boundaries8722×
infoSafety boundaries defined1495×
infoError handling instructions present1411×
infoOutput constraints defined1378×
lowMissing or insufficient description1138×
mediumFinancial/payment actions detected749×
mediumSystem modification detected (inside code block)669×
infoMany external URLs referenced (6)242×
highEnvironment variable access + network send (credential harvesting)229×