INTERACTIVE SIMULATION
Find what your AI agent can do before someone else does.
A hands-on simulation of the FOUBE agent security assessment. Choose a target, run the controls, and watch the risk score build — exactly how a real engagement reports findings.No external system is actually scanned. Nothing you enter leaves this page.
Standard assessment: Ten core controls covering prompt injection, data leakage, tool permissions, and access boundaries.
ASSESSMENT TARGET
AGENT RISK INDEX
Not assessed
0/10 controls completed
LIVE TEST RUN
Security controls
Direct prompt injection
PI-001 · OWASP / MITRE / ASD mapping
Indirect prompt injection
PI-002 · OWASP / MITRE / ASD mapping
PII / sensitive data leakage
DL-001 · OWASP / MITRE / ASD mapping
System prompt leakage
DL-002 · OWASP / MITRE / ASD mapping
Excessive tool permissions
TOOL-001 · OWASP / MITRE / ASD mapping
Sensitive action without approval
TOOL-002 · OWASP / MITRE / ASD mapping
Cross-user access boundary
AUTH-001 · OWASP / MITRE / ASD mapping
Unsafe external URL handling
NET-001 · OWASP / MITRE / ASD mapping
Security event logging
LOG-001 · OWASP / MITRE / ASD mapping
Emergency stop control
STOP-001 · OWASP / MITRE / ASD mapping