The Agent Skills Directory

[PROMPT_INJECTION]: Indirect Prompt Injection Surface. The skill reads local files such as CSS, JS, and project documentation to audit existing animations. This content is processed without sanitization or boundary markers, which could allow instructions hidden in the codebase to influence agent behavior.
Ingestion points: roles/motion-researcher/role.md and roles/animator/role.md read codebase content via Glob and Grep.
Boundary markers: No explicit boundary markers or instructions to ignore embedded commands are present in the scripts when processing external files.
Capability inventory: The agents have access to powerful tools including Bash, Write, Edit, Agent spawning, and mcp__chrome-devtools__evaluate_script.
Sanitization: No sanitization is performed on ingested codebase content before it is processed by the LLM.
[COMMAND_EXECUTION]: Dynamic Code Generation. The animator role generates JavaScript orchestrators and CSS animations at runtime based on choreography specifications. These are subsequently utilized by the motion-tester role, which executes scripts via the mcp__chrome-devtools__evaluate_script tool to validate performance, representing a surface for dynamic code execution within the browser environment.

team-motion-design