The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes kubectl, okteto, jq, and cat commands to inspect the status of the environment, retrieve logs, and validate configuration files. These operations are diagnostic in nature and do not perform unauthorized modifications.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface because it ingests untrusted application logs and cluster event data.
Ingestion points: SKILL.md (via commands kubectl logs, okteto logs, kubectl describe, and kubectl get events).
Boundary markers: None specified in the instructions to delimit or isolate log content.
Capability inventory: SKILL.md (invokes okteto context, cat, kubectl, okteto status, okteto logs, and okteto validate).
Sanitization: No sanitization or escaping of the ingested log output is performed.

okteto-debugging