The Agent Skills Directory

[COMMAND_EXECUTION]: Executes standard kubectl commands to inspect pod statuses, configurations, and init container specifications to verify successful APM instrumentation.
[COMMAND_EXECUTION]: Utilizes the pup CLI tool (a Datadog-related utility) to verify authentication and check for the presence of services and traces via the Datadog API.
[DATA_EXFILTRATION]: Constructs deep links to the Datadog platform (app.<DD_SITE>) to allow users to navigate directly to their service metrics. This activity is restricted to the vendor's known domains and is consistent with the skill's primary purpose.
[PROMPT_INJECTION]: The skill has an indirect prompt injection surface as it processes external data from command outputs.
Ingestion points: Command line output from kubectl and pup (SKILL.md).
Boundary markers: None; the agent is instructed to fill the report table directly with the command outputs.
Capability inventory: The skill is limited to read-only status and diagnostic commands.
Sanitization: The skill assumes the integrity of the data returned by the cluster and the vendor API.

onboarding-summary