design

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

• Third-party content exposure detected (medium risk: 0.65). The required runtime workflow reads multiple outsider-authored research artifacts from artifacts/output/01-research/* (e.g., market-analysis.md, competitive-analysis.md, user-personas.md), which are typically populated from public web content or other third-party sources during the prior exploration step, and those readable markdown contents are ingested into the LLM context by @product-manager/@product-designer (“Context budget: PM reads discovery brief in full, then reads only the executive summary + relevant sections of each research artifact”).