Understanding your score

Score ranges

Score	Label	Meaning
90–100	Passed	No significant issues — cluster is well hardened
70–89	Low Risk	Minor issues found, no immediate action required
50–69	Medium Risk	Several issues detected — review recommended soon
30–49	High Risk	Significant vulnerabilities found — action required
0–29	Critical	Cluster has critical security issues — act immediately

The formula

The score uses a diminishing-returns penalty model — fixing your first critical finding has more impact than fixing your 10th.

penalty(count, perIssue, cap) = min(cap, perIssue × (1 − 0.75ⁿ) / 0.25)

score = 100
      − penalty(critical, 18, 42)
      − penalty(high,     10, 28)
      − penalty(medium,    4, 14)
      − penalty(low,       1,  6)

Max deduction per severity

Severity	Per issue	Max deduction
Critical	−18 pts	−42 pts
High	−10 pts	−28 pts
Medium	−4 pts	−14 pts
Low	−1 pt	−6 pts

Example scores

Findings	Score
0 of everything	100
1 critical	72
4 critical, 30 high, 6 medium, 31 low	13
2 high, 5 medium	73

Why does my fresh minikube score so low?

minikube is not hardened by default. A typical scan finds:

• kube-proxy running with hostNetwork: true → critical
• Missing NetworkPolicies on all namespaces → high
• Containers with allowPrivilegeEscalation not set to false → high (one per container)
• Images using :latest tag → medium
• Missing resource limits → medium

This is expected — minikube is a local dev tool, not a production cluster. On a properly hardened EKS cluster you'd typically score 70+.

Improving your score

The Findings tab shows every issue with a remediation guide. Sort by severity and work top-down. Fixing 1 critical finding typically adds 15–20 points.