databricks-cluster-forensics

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). Skill databricks-cluster-forensics (via the cluster-event-investigator subagent and scripts/cluster-coldstart-forensics.py) ingests Databricks clusters_events/termination payloads (including parameters.databricks_error_message free text) that an outsider can influence by triggering/causing events in a cluster they control, e.g., in advisory mode where it “accepts pasted clusters.events JSON”.