k8s-autoscaling
Fail
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: HIGHCREDENTIALS_UNSAFECOMMAND_EXECUTION
Full Analysis
- [CREDENTIALS_UNSAFE] (HIGH): Hardcoded credential pattern detected in example configuration file.
- Evidence: Found
amqp://user:pass@rabbitmq.default.svc:5672inKEDA-TRIGGERS.md. While used in an example, such patterns can be mistakenly applied to production environments or leak placeholder formats. - [COMMAND_EXECUTION] (HIGH): High-privilege capability to modify Kubernetes cluster state via manifest application (Category 8: Indirect Prompt Injection surface).
- Ingestion points: The skill processes YAML definitions provided in the documentation and examples (e.g.,
hpa-cpu.yaml,keda-scaledobject.yaml). - Boundary markers: None identified. There are no explicit instructions to the agent to validate or sanitize the content of the YAML before application.
- Capability inventory: The tool
kubectl_applyis used inSKILL.mdto commit arbitrary resource definitions to the cluster. - Sanitization: Absent. The agent is encouraged to 'apply and verify' without manual validation steps.
Recommendations
- AI detected serious security threats
Audit Metadata