scientific-critical-thinking

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

• Potentially malicious external URL detected (high risk: 0.90). The skill includes runtime calls to the OpenRouter API (https://openrouter.ai/api/v1) — the scripts require an OPENROUTER_API_KEY and use model responses (Gemini image generation and Gemini 3.1 quality review) to drive iterative diagram generation and decision-making, so external content directly controls the agent's behavior.