3d-generator
Pass
Audited by Gen Agent Trust Hub on Aug 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data including text prompts and image URLs. These are transmitted to external 3D generation services via an API gateway, representing an indirect prompt injection surface. * Ingestion points: The 'prompt', 'images', and 'multi_view' fields within the task creation API examples in SKILL.md. * Boundary markers: Not present in the provided interaction examples to delimit user data from instructions. * Capability inventory: Performs network requests to agent.deepnlp.org and handles model download URLs. * Sanitization: No explicit validation or sanitization of user-provided content is documented.
- [EXTERNAL_DOWNLOADS]: The skill documentation includes examples using 'npx onekey agent', which initiates runtime downloading and execution of the 'onekey' package from the npm registry.
- [COMMAND_EXECUTION]: The skill relies on shell command execution (curl, npx) for its core functionality and interaction examples, which the agent is instructed to perform.
Audit Metadata