CC auditing-kubernetes-rbac-privilege-escalation
Finds over-permissive RBAC roles and service-account token abuse paths in a Kubernetes cluster using kubectl auth can-i, rbac-police, kubectl-who-can, and rakkess, tracing which subjects can escalate toward cluster-admin. Use when reviewing who can escalate privileges in a cluster, hunting exploitable RoleBindings during an authorized review, or validating least privilege after an RBAC change. Keywords: RBAC, ClusterRoleBinding, service account token, auth can-i, rbac-police, escalate, bind, impersonate. Do not use for designing and applying hardened RBAC - use implementing-rbac-hardening-for-kubernetes.
Finds over-permissive RBAC roles and service-account token abuse paths in a Kubernetes cluster using kubectl auth can-i, rbac-police, kubectl-who-can, and…
As a process C 52/100 · Has gaps — weak spots: result and completion, failures and branches, running it twice
How to improve
- Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
- A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.
Guard findings · 9
✓ No critical or high findings
Medium and low: 9
-
low Risky intent
intent-offensive-securityreferences/standards.md:1Offensive-security / dual-use content (legitimate for authorised testing; review intended use)# Standards and References - Kubernetes RBAC Privilege Escalation Audit
-
low Risky intent
intent-offensive-securityreferences/standards.md:7Offensive-security / dual-use content (legitimate for authorised testing; review intended use)| T1078 | Valid Accounts | Defense Evasion / Privilege Escalation | RBAC abuse leverages legitimate service-account credentials to gain higher access. |
-
low Risky intent
intent-offensive-securityreferences/standards.md:11Offensive-security / dual-use content (legitimate for authorised testing; review intended use)| T1611 | Escape to Host | Privilege Escalation | `create pods` plus node access yields a privileged host-mounting pod. |
-
low Risky intent
intent-offensive-securityreferences/standards.md:31Offensive-security / dual-use content (legitimate for authorised testing; review intended use)- Unit 42: Kubernetes RBAC privilege escalation research
-
low Risky intent
intent-offensive-securitySKILL.md:30Offensive-security / dual-use content (legitimate for authorised testing; review intended use)# Auditing Kubernetes RBAC Privilege Escalation
-
low Risky intent
intent-offensive-securitySKILL.md:54Offensive-security / dual-use content (legitimate for authorised testing; review intended use)- During an authorized Kubernetes security assessment or cluster penetration test
-
low Risky intent
intent-offensive-securitySKILL.md:92Offensive-security / dual-use content (legitimate for authorised testing; review intended use)| T1078 | Valid Accounts | Defense Evasion / Persistence / Privilege Escalation |
-
low Risky intent
intent-offensive-securitySKILL.md:96Offensive-security / dual-use content (legitimate for authorised testing; review intended use)| T1611 | Escape to Host | Privilege Escalation |
A further 1 matches are quotations in this security skill's documentation and are not counted as findings.
Files scanned: 5. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- note
frontmatter-keyunknown frontmatter key "domain" - note
frontmatter-keyunknown frontmatter key "subdomain" - note
frontmatter-keyunknown frontmatter key "nist_csf" - note
frontmatter-keyunknown frontmatter key "mitre_attack"
Process rating: all ten parameters 52/100
- 0Result and completion. Does not say what the result is
- 0Failures and branches. Linear process with no failure handling
- 0Progress reporting. Says nothing while it works
- 30Running it twice. 9 mutating operations with no state check
- 50When it triggers. No condition that starts the skill
- 60Tools and files. Uses tools (node) that frontmatter does not declare
- 70Inputs and preconditions. Inputs and preconditions are listed
- 85Steps. 31 steps, 1 vague phrases
- 100Consistency. Name and required fields are in place
- 100Execution cost. Instruction body is 2779 tokens
Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.
Quality signals
- +5Description has no quoted example phrases that should trigger the skill
- +3Output format is not stated: the model decides each time
- -42 reference files, but SKILL.md never points to them: the model will not open them
- -31 of 1 scripts are never mentioned in SKILL.md
- +2Single-language instructions
- +4Description says when NOT to use the skill
- +3Description length 611: enough signal without eating the budget
- +4Structure: 17 headings
- +3Step-by-step instructions: 31 items
- +4Has examples (8 code blocks)
- +1License stated
Quality base 70; lint remarks subtract, signals add up to 100. Result: 79.