k8s-manifest-generator
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied application requirements to generate Kubernetes manifests. It lacks specific boundary markers or sanitization logic to prevent malicious input from being interpolated into the generated YAML files.
- Ingestion points: Requirement gathering phase described in
SKILL.md, where the agent collects workload details from the user. - Boundary markers: Absent; there are no explicit delimiters or instructions to ignore potential instructions embedded within the user's data.
- Capability inventory: The generated manifests can include execution-related configurations such as init containers, lifecycle hooks, and shell scripts in ConfigMaps (
references/deployment-spec.md,assets/configmap-template.yaml). - Sanitization: No validation or escaping is applied to user-provided values before they are placed in the manifest templates.
Audit Metadata