The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests external data from monitoring backends that could contain attacker-controlled content. Ingestion points: Alert data and health statuses are fetched from Datadog, New Relic, and OpenTelemetry endpoints. Boundary markers: The instructions do not define clear boundaries or warnings for the subagent processing this external data. Capability inventory: The skill utilizes Bash and Agent tools, providing a wide range of system capabilities. Sanitization: The skill does not explicitly sanitize the text retrieved from API responses before displaying it or passing it to subagents.
[EXTERNAL_DOWNLOADS]: The skill makes network requests to well-known service providers (Datadog and New Relic) to fetch operational data. These requests are directed at established, official domains.
[COMMAND_EXECUTION]: The skill employs bash commands for routine tasks such as configuration management via jq and network polling via curl. These operations are consistent with the skill's stated monitoring purpose.

ops-monitor