prd-v01-user-value-articulation

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

• Third-party content exposure detected (high risk: 0.90). The skill's references/research-prompts.md explicitly instructs the agent to search and extract evidence from public sources (e.g., Reddit/forums, G2/Capterra, competitor sites, job postings) so the agent would ingest untrusted, user-generated web content that can change evidence tiers and downstream decisions.