The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from Datadog logs, traces, and incidents, which could contain malicious instructions.
Ingestion points: Data is ingested through pup CLI commands such as logs search, traces search, and incidents get (documented in references/api-surface.md).
Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions when interpolating retrieved telemetry into the agent context.
Capability inventory: The skill has the capability to execute shell commands using the pup tool and can read/write to the terminal output (documented in SKILL.md).
Sanitization: No technical sanitization or validation of the content retrieved from Datadog is described beyond the instruction to quote only the minimum content needed.

datadog