gcp-playbooks

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

• Third-party content exposure detected (medium risk: 0.30). At runtime, the skill’s script scripts/report_errors_by_revision.py fetches Cloud Error Reporting representative.message/stack traces from GCP APIs (via service.projects().events().list(...)) and feeds that free-text into the output (LLM ingestion risk if an agent reads/uses it), which originates from outsider-authored application error content.