hugging-face-evaluation

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

• Third-party content exposure detected (medium risk: 0.30). In the extract-readme CLI workflow (used via inspect-tables then extract-readme --repo-id ... --table N), the agent ingests and parses model-card README markdown from a Hugging Face repo—i.e., outsider-authored free text that can be attacker-supplied by posting to that README.