form-style

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

• Third-party content exposure detected (high risk: 0.70). The workflow runs python3 -m form_agent.uiux search for both product and style domains, which at runtime likely fetches/searches external (public/community) content and feeds the resulting free-text fields (e.g., “Anti_Patterns”, “key_effects”) into the agent’s LLM context via those search results.