The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes Cinema 4D scene data (object names, material names, and node graph properties), which constitutes a potential attack surface for indirect prompt injection if those fields contain adversarial instructions.
Ingestion points: Untrusted data enters the context through tools like list_objects, get_scene_info, and the custom Redshift material inspection patterns which read user-defined names from the scene.
Boundary markers: Absent. The instructions do not recommend using delimiters or specific 'ignore' instructions when processing strings extracted from scene objects.
Capability inventory: The skill leverages execute_python_script to run Python code within Cinema 4D and uses standard library modules (os, tempfile) for local data management.
Sanitization: Absent. The skill does not provide methods for sanitizing scene metadata before it is interpolated into the agent's reasoning or tool parameters.

cinema4d-mcp