omniverse-cad-to-simready

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

• Third-party content exposure detected (high risk: 0.75). Outsider free text can enter the LLM context via identify-asset-context web research: it performs runtime web searches and writes asset-context.md/json evidence and prompts that are then passed as --prompt "$ASSET_CONTEXT_PROMPT" into Content Agents (material/physics/texture), so third-party web page text can be included in the prompt sent to the agents’ LLM.