The Agent Skills Directory

[DYNAMIC_EXECUTION]: The checker script dynamically imports TypeScript modules from the project being analyzed to evaluate the animation catalog.
Ingestion point: scripts/lib/app-modules.mjs uses the import() function with dynamically constructed URLs to load files from the audited project (e.g., src/ui/motion/motion-moments.ts).
Purpose: Loads the application's animation metadata to perform validation against project-specific constants.
Mechanism: Utilizes Node.js module.registerHooks to intercept and transform module resolution for TypeScript support.
[INDIRECT_PROMPT_INJECTION]: The skill creates an attack surface by ingesting and printing strings from untrusted application source files back to the agent's context.
Ingestion points: scripts/check-motion.mjs and scripts/lib/source-scan.mjs walk the audited repository and read .ts and .tsx files.
Boundary markers: Extracted "trigger" descriptions and moment metadata (which are arbitrary strings in the user's code) are printed in tables or logs without delimiters or "ignore instructions" warnings.
Capability inventory: scripts/check-motion.mjs performs extensive file reading and outputs extracted content to the console, which is subsequently interpreted by the AI agent.
Sanitization: No filtering or escaping is performed on strings extracted from the audited codebase before they are presented to the agent.

motion-and-animation