Skip to content

Security Posture

The security posture score (0-100) provides a comprehensive view of your AI agent's security health. The formula is:

posture = 100 * (1 - severity_impact) * coverage_effectiveness
  • Severity Impact: Weighted sum of active findings (by severity and status)
  • Coverage Effectiveness: Ratio of tested threat classes with acceptable pass rates

View Current Posture

hb posture

hb posture --trends

Export as JSON

hb posture --json

View Specific Project

hb posture --project <id>

Organisation-Level Posture (3 Dimensions)

# Aggregate posture across all projects + inventory
hb posture --org

Org posture aggregates three dimensions:

  • Agent Security -- posture across all security-tested projects
  • Shadow AI -- risk posture of discovered AI inventory assets
  • Quality -- behavioral quality across all tested agents

Include Coverage Breakdown

hb posture --coverage

Posture Grades

Grade Range Description
A 90-100 Excellent security posture with minimal vulnerabilities and comprehensive coverage
B 75-89 Good security posture with minor vulnerabilities or coverage gaps
C 60-74 Fair security posture requiring attention to identified issues
D 40-59 Poor security posture with significant vulnerabilities or coverage gaps
F 0-39 Critical security posture requiring immediate remediation