The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes data from external Figma nodes (such as keyframes, animation styles, and timelines), creating an attack surface where malicious instructions could be embedded in design metadata.
Ingestion points: Node property reads documented in references/motion-patterns.md (e.g., manualKeyframeTracks, animationStyles).
Boundary markers: The skill does not use explicit delimiters or instructions to ignore embedded prompts within the ingested Figma data.
Capability inventory: The skill can modify design nodes and initiate server-side video renders via the use_figma tool.
Sanitization: No sanitization or validation of the ingested Figma data is mentioned before its use in subsequent logic.
[DYNAMIC_EXECUTION]: The skill provides templates for the generation of code and shell commands that are executed at runtime to interact with the Figma API and local system utilities.
Evidence: Numerous JavaScript code templates in references/motion-patterns.md and references/motion-easing.md designed for execution through the use_figma tool.
Evidence: Shell command patterns for ffmpeg frame extraction documented in the SKILL.md file.

figma-use-motion