The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes data from PagerDuty (such as incident titles, notes, and alerts) which can contain malicious instructions from external actors.
Ingestion points: Data enters via tools like PAGERDUTY_FETCH_INCIDENT_LIST, PAGERDUTY_RETRIEVE_INCIDENT_BY_INCIDENT_ID, PAGERDUTY_GET_ALERTS_BY_INCIDENT_ID, and PAGERDUTY_GET_SCHEDULES.
Boundary markers: The instructions do not specify any delimiters or warnings to help the agent distinguish between trusted instructions and untrusted data content.
Capability inventory: The skill possesses extensive write capabilities, including creating/updating incidents, services, schedules, and escalation policies.
Sanitization: There are no instructions for the agent to sanitize, escape, or validate the content retrieved from PagerDuty before using it in decision-making or workflows.
[EXTERNAL_DOWNLOADS]: The skill requires the user to configure a remote MCP server at https://rube.app/mcp. Connecting to an external endpoint allows a third party to provide tool definitions and execute logic within the agent's runtime environment.

pagerduty-automation