hypothesis-generation

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

• Third-party content exposure detected (medium risk: 0.30). The runtime path reads outsider-authored free text directly from the user-provided diagram description (prompt) and sends it to the LLM in scripts/generate_schematic.py → scripts/generate_schematic_ai.py (generate_iterative(user_prompt=...) → generate_image() → _make_request() with messages=[{"role":"user","content": prompt}]).

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 1.00). The scripts call the OpenRouter API at https://openrouter.ai/api/v1 at runtime to generate images and receive review critiques that directly drive prompt updates and the iterative generation flow (requires OPENROUTER_API_KEY), so this external endpoint controls agent behavior.