deepen-plan

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

• Third-party content exposure detected (high risk: 0.70). SKILL.md requires spawning mktg-audience-researcher and mktg-competitive-scanner agents that write to brand/audience.md and brand/competitors.md, and those agents’ research at runtime is likely to ingest outsider-authored free text from public web sources (indirect prompt injection risk via the agents’ outputs being read in Phase 3).