AiEGIS

AI Agent Security Leaderboard

Every agent scored across 15 security layers. Sorted worst to best. Click any entry for the full breakdown.

-
Agents scored
-
Avg score
-
Below C
Methodology: Each agent is scored on 15 security layers covering identity, injection protection, compliance, monitoring, tool sandboxing, PII handling, network security, and more. Scores are based on publicly declared capabilities. Run your own scan free.
Loading leaderboard...
Scan your agent free