argocd-ops
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONPRIVILEGE_ESCALATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes various shell commands to interact with the ArgoCD CLI and Kubernetes API. This includes using python3 for parsing JSON data and running argocd CLI commands.
- [PRIVILEGE_ESCALATION]: The skill performs administrative tasks that require high-level access to the Kubernetes cluster. It uses kubectl patch to modify resource configurations, rollout restart to cycle deployments, and kubectl exec to run commands inside infrastructure pods such as argocd-server and argocd-redis. Additionally, the environment configuration allows for sudo to be embedded in the KUBECTL command.
- [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection.
- Ingestion points: Untrusted data enters the agent's context through kubectl get application -ojson and the output of argocd app diff (SKILL.md).
- Boundary markers: There are no explicit delimiters or instructions to the agent to disregard potential instructions embedded within the Kubernetes resource data or diff results.
- Capability inventory: The skill possesses significant capabilities, including writing to the cluster via kubectl patch, performing kubectl exec into pods, and restarting deployments (SKILL.md).
- Sanitization: No sanitization or validation of the ingested Kubernetes resource data is implemented before the agent interprets the content.
Audit Metadata