configuring-service-meshes
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes shell scripts (deploy_config.sh, rollback_config.sh) and a Python script (generate_test_traffic.py) to execute kubectl commands. These operations are core to the skill's functionality for managing Kubernetes resources and are restricted to the context of service mesh administration.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied Kubernetes manifests and service URLs, which represents a potential attack surface for indirect prompt injection.
- Ingestion points: Kubernetes YAML manifest files processed by validate_config.py and deploy_config.sh; service URLs provided to generate_test_traffic.py.
- Boundary markers: The skill relies on standard Kubernetes resource delimiters and does not implement specific 'ignore' instructions for embedded data.
- Capability inventory: The skill can create, update, and delete Kubernetes resources (Deployments, Services, Istio CRDs) and initiate network requests via a temporary pod running curl.
- Sanitization: The validate_config.py script correctly uses yaml.safe_load_all() to mitigate risks associated with unsafe YAML deserialization.
Audit Metadata