Skills
skills/wgpsec/aboutsecurity/k8s-webhook-abuse/Gen Agent Trust Hub

k8s-webhook-abuse

Fail

Audited by Gen Agent Trust Hub on Apr 22, 2026

Risk Level: HIGHDATA_EXFILTRATIONCOMMAND_EXECUTIONCREDENTIALS_UNSAFE
Full Analysis
  • [DATA_EXFILTRATION]: The primary purpose of the skill is to exfiltrate sensitive data, such as secrets and flags, from Kubernetes Mutating Webhooks by spoofing internal API Server requests.
  • Evidence: Phase 3 in SKILL.md describes extracting and decoding the response.patch field from a forged Webhook response to reveal injected environment variables and secrets.
  • [COMMAND_EXECUTION]: The skill utilizes command-line tools to perform network reconnaissance and active exploitation within a Kubernetes environment.
  • Evidence: SKILL.md instructs the use of k8spider scan for service discovery and curl with the -k (insecure) flag to POST forged JSON payloads to internal service endpoints.
  • [CREDENTIALS_UNSAFE]: The described workflow is explicitly designed to harvest credentials and sensitive configuration data (tokens, secrets) that are dynamically injected into Kubernetes resources.
  • Evidence: The skill description and Phase 2 outline techniques for 'tricking' the Webhook into returning patches intended for privileged Pods, containing secrets or flags.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Apr 22, 2026, 10:08 AM