prd-v01-problem-framing

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

• Third-party content exposure detected (high risk: 1.00). The skill's required research workflow (see references/research-prompts.md and SKILL.md Step 1/Step 2) explicitly instructs searching and extracting verbatim quotes from open/public user-generated sources such as Reddit/forums, G2/Capterra reviews, job boards, and support forums, meaning the agent will ingest untrusted third‑party content that can influence decisions.