recent-discourse-sweep

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

• Third-party content exposure detected (medium risk: 0.30). The required runtime workflow explicitly instructs the agent to “list the venues before searching” and then “run the venues one at a time” including “open discussion communities… short-form social feeds… plain web search” and “scraped post, comment, listing or transcript,” which means outsider-authored free text is ingested during the sweep.