dev-swarm-stage-tech-research

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

• Third-party content exposure detected (high risk: 0.80). The skill's Research Execution Rules explicitly require fetching "latest docs from internet" for third-party APIs (see "Third-party APIs: Fetch latest docs from internet and save key information in the research folder"), meaning the agent will ingest public third‑party web content that directly informs research decisions and implementations and could therefore enable indirect prompt injection.