The Agent Skills Directory

[COMMAND_EXECUTION]: The skill employs system binaries and project-specific CLI tools to manage design assets and verify animation fidelity.\n
Execution of ffprobe and ffmpeg within scripts/verify-motion.mjs for motion energy analysis.\n
Use of hyperframes CLI commands for asset, token, and component synchronization.\n- [EXTERNAL_DOWNLOADS]: The skill provides instructions for maintaining its components via external updates.\n
npx hyperframes skills update figma is used to fetch and update skill dependencies from the vendor's repository.\n- [DYNAMIC_EXECUTION]: The skill generates executable content at runtime based on ingested design data.\n
Design motion context is normalized and converted into GSAP timeline scripts, which are then injected into the composition environment.\n
Local execution of verification scripts via node is part of the standard workflow.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes external design data from Figma, presenting a surface for indirect instructions.\n
Ingestion points: Figma REST API responses and connector-provided design trees (referenced in SKILL.md).\n
Boundary markers: None explicitly mentioned for isolating ingested text or metadata from the instruction flow.\n
Capability inventory: Command execution, file system writes to .media/, and script injection into the final composition.\n
Sanitization: The process relies on structural normalization without documented safety delimiters or content filtering for design-originated strings.

hyperframes-figma