The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized access attempts were detected in the skill instructions. The skill focuses on providing documentation and examples for using 3D generation tools via a legitimate platform API.
[EXTERNAL_DOWNLOADS]: The skill recommends installing sibling skills from the vendor's own repository (scenario-labs/skills) via npx if they are missing. This is a standard procedure for managing dependencies in the supported environment and targets the author's official resources.
[INDIRECT_PROMPT_INJECTION]: The skill defines an interface that accepts user-provided text prompts and external assets (images and 3D models) as input for the Rodin generation models. 1. Ingestion points: prompt, images, and model parameters in model_run (SKILL.md). 2. Boundary markers: The parameters are passed within a structured JSON object (parameters={...}) in the model_run tool call. 3. Capability inventory: Subprocess-like capability via the model_run and upload_asset tools (SKILL.md). 4. Sanitization: Not explicitly handled within the skill instructions; relies on the underlying platform's safety filters.

scenario-rodin