The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data, creating a surface for potential indirect prompt injection attacks.
Ingestion points: The workflow instructs the agent to analyze error logs (via kubectl logs), health check responses (via curl), and monitoring metrics from external systems (Prometheus/Grafana) to perform "Incident Translation" and "Operational Research".
Boundary markers: The skill does not define specific delimiters or instructions for the agent to treat content from logs or network outputs as untrusted data to be ignored as instructions.
Capability inventory: The agent possesses powerful capabilities including Kubernetes cluster management (kubectl), container lifecycle control (docker), and database interactions (psql).
Sanitization: While the workflow incorporates pre-deployment security scanning tools like Snyk and npm audit, it lacks explicit runtime sanitization or filtering for data retrieved during incident analysis or monitoring steps.

devops-engineer