The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes dynamic context injection to run shell commands at the time the skill is loaded. Specifically, it executes kubectl config current-context, which retrieves and inserts the active Kubernetes infrastructure context into the agent's environment without user confirmation or oversight.
[PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted external data including symptoms, error messages, system logs, and metrics. This data is used to drive automated investigation tracks, creating a potential surface for indirect prompt injection. * Ingestion points: Affected systems, logs, metrics, and alerts defined in the Input section. * Boundary markers: None identified. * Capability inventory: The skill launches multiple agents in parallel for external research and codebase exploration. * Sanitization: There is no evidence of filtering or escaping logic for the provided input data.

incident