create-scene

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

• Third-party content exposure detected (high risk: 0.70). The SKILL.md explicitly instructs the agent to find and download models from public URLs (e.g., the curl example fetching from https://builder-items.decentraland.org/...) and to select/use those third‑party assets as part of the scene-building workflow, which exposes the agent to untrusted external content that can influence its decisions and actions.