markbuildbrand-foundation-docs

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

• Third-party content exposure detected (high risk: 1.00). This skill's required workflow (Step 1, Prompt #4 in SKILL.md and the "Research Sources" list in reference/deep-research-methodology.md) explicitly instructs the agent to run a "Deep Research" tool that fetches and copies research from public, user-generated sources (Amazon reviews, Reddit/forums, social media, Google results) into docs that will be used to drive subsequent copywriting decisions, exposing the agent to untrusted third-party content that can influence actions.