The Agent Skills Directory

[COMMAND_EXECUTION]: The skill contains numerous templates featuring powerful operational shell commands. It includes Kubernetes management tasks using kubectl for scaling, rollbacks, and log inspection; PostgreSQL operations via psql for connection management and process termination; and network requests using curl to interact with health checks and internal feature flag APIs. These commands are consistent with the skill's purpose as an incident response tool.\n- [EXTERNAL_DOWNLOADS]: The templates reference several well-known services and status monitoring platforms, including Grafana, Sentry, PagerDuty, and Stripe. These are documented as standard external resources for monitoring and verifying service status during active incidents.\n- [PROMPT_INJECTION]: The skill exhibits a potential surface for indirect prompt injection.\n
Ingestion points: The templates instruct the agent to retrieve and analyze external data streams, specifically system logs (kubectl logs) and Prometheus metrics.\n
Boundary markers: The provided templates do not specify the use of delimiters or instructions to ignore embedded commands within the ingested log content.\n
Capability inventory: The skill leverages sensitive tools including kubectl, psql, curl, and local shell script execution.\n
Sanitization: No explicit sanitization or validation logic is defined for the external content processed by the agent.

incident-runbook-templates