research-brainstorm

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

• Third-party content exposure detected (high risk: 0.90). The skill's required Phase 4 "Parallel novelty scan" explicitly instructs launching deep literature scans across Google Scholar, working-paper servers, and "visible econ social-media chatter" (and Phase 8 mentions web-scraped public repositories), i.e., fetching and ingesting open/public third‑party content that the agent is expected to read and that materially influences its decisions and tool use.