brand-guidelines

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

• Third-party content exposure detected (high risk: 0.80). The SKILL.md Brand Audit workflow explicitly requires "Collect samples from every active channel" including website pages and social media posts (Workflow 2: Step 1), and the included scripts (messaging_consistency_checker.py, brand_audit_scorer.py) expect the agent to ingest and analyze those external, user-generated web/social samples — meaning untrusted third‑party content is read and can materially influence recommendations and next actions.