owasp-kubernetes
Pass
Audited by Gen Agent Trust Hub on Aug 25, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates an audit workflow that requires the agent to ingest and evaluate external, potentially untrusted Kubernetes manifests, logs, and configurations.
- Ingestion points: The
SKILL.mdentrypoint defines a required assessment workflow where the agent must 'Establish the assessment scope, evidence sources, and relevant Kubernetes resources' for analysis. - Boundary markers: The instructions do not define or enforce the use of protective delimiters (e.g., XML tags or specific separators) or include explicit warnings for the agent to ignore instructions embedded within the audit evidence.
- Capability inventory: The agent is expected to use its available toolset to read files and potentially query cluster APIs to retrieve the resources defined in the scope, allowing external data to enter the agent's context.
- Sanitization: There are no documented steps for the agent to validate, escape, or sanitize the external data before it is interpolated into the analysis prompt, which could allow a malicious Kubernetes manifest to influence the auditor's findings or logic.
Audit Metadata