capy-cortex

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

• Potentially malicious external URL detected (high risk: 0.90). The skill calls external LLM endpoints at runtime via CORTEX_API_URL (e.g. https://openrouter.ai/api/v1) and uses the model outputs for causal analysis, quality gating, and to generate/store rules that are injected into agent context, meaning remote responses can directly control agent prompts/instructions.