launch-monitor

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

• Third-party content exposure detected (high risk: 0.75). At runtime the skill ingests outsider-authored free text from public/community sources via connectors—e.g., scripts/connectors/hn.py (HN comments) and scripts/connectors/gdelt.py (news echo)—and treats “every API response” as untrusted input, which can be fed into the agent’s LLM context for analysis/alerting.