The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the execute_blender_code tool to run Python commands within a Blender session, which is the intended method for automating animation and rigging tasks.
[DYNAMIC_EXECUTION]: The skill instructs the agent to generate and execute Python scripts for complex tasks such as armature creation and driver setup, utilizing standard Blender API patterns.
[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it reads data from the Blender scene and API documentation to inform its code generation.
Ingestion points: Untrusted data enters the context through scene summaries (get_objects_summary) and search results (search_api_docs) in SKILL.md.
Boundary markers: No specific delimiters or warnings are provided to prevent the agent from following instructions potentially embedded in scene data.
Capability inventory: The skill utilizes execute_blender_code for live session mutation, along with various UI manipulation and rendering tools across all documentation files.
Sanitization: The skill lacks explicit sanitization steps for scene-derived data before it is interpolated into Python scripts.

blender-animation-rigging