k8s-flux-troubleshooting
Pass
Audited by Gen Agent Trust Hub on Nov 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data retrieved from the Kubernetes cluster, including container logs, resource event messages, and object descriptions. This data can be influenced by an external attacker (e.g., through a compromised Git repository or unauthorized cluster access), potentially embedding instructions intended to manipulate the agent.
- Ingestion points: The skill utilizes
kubectl logs,kubectl events, andkubectl describeacrossSKILL.mdand all files in thereferences/directory to inspect cluster state. - Boundary markers: There are no explicit delimiters or instructions provided to the agent to treat the output of these diagnostic commands as untrusted data or to ignore instructions contained within them.
- Capability inventory: The skill is capable of performing cluster modifications via
kubectl patchandkubectl annotate(e.g., in the 'Proposed fix' sections ofSKILL.mdandreferences/helmreleases.md), which could be targeted by an injection attack. - Sanitization: The skill does not perform validation or sanitization of the cluster data before processing it.
Audit Metadata