The Agent Skills Directory

[Indirect Prompt Injection] (LOW): The skill possesses an indirect prompt injection surface due to ingestion of untrusted data from the cluster environment.\n
Ingestion points: The skill uses argocd app logs (in SKILL.md and TroubleshootSync.md) and argocd app manifests (in ReviewResources.md) which retrieve potentially attacker-controlled content from the managed cluster into the agent's context.\n
Boundary markers: Absent. The documentation and workflows do not include any delimiters or instructions for the agent to ignore instructions embedded in the logs or manifests.\n
Capability inventory: The skill provides commands with high impact on infrastructure, including argocd app sync, argocd app rollback, and argocd app set (found in RollbackApp.md and TroubleshootSync.md).\n
Sanitization: Absent. There is no mechanism described for sanitizing or validating the retrieved data before processing.\n- [Command Execution] (SAFE): The commands provided for argocd and kubectl are standard for the tool's intended use case. No obfuscation, persistence mechanisms, or unauthorized privilege escalation attempts were found.

ArgoCDReview