community-pulse

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly fetches and reads user-generated Discord content (e.g., get_channel_messages in Step 5 and the Topic analysis in Step 7) from a third-party public platform and uses that content to generate topics, quotes, and action recommendations, so untrusted messages can materially influence the agent's decisions.