blender-mcp

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

• Third-party content exposure detected (high risk: 1.00). The skill's Asset Integrations (SKILL.md and references/asset-integrations.md) explicitly fetch and auto-import public third‑party assets from PolyHaven and Sketchfab (and cloud generators like Hyper3D Rodin), then the workflow scripts/inspections (execute_python examples) read and act on those imported materials/objects (e.g., baking, material patches, export decisions), so untrusted external content is ingested and can materially influence subsequent tool actions.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill repeatedly invokes the gltf-transform CLI via npx (e.g., "npx @gltf-transform/cli ..."), which will fetch and execute remote npm package code at runtime from https://www.npmjs.com/package/@gltf-transform/cli and is a required dependency for the optimization pipeline.