mt-create-skill

To create a reusable AI agent workflow, you need a structured definition file containing metadata, operational steps, completion criteria, and constraints. A guided workflow collects these requirements and validates the structure to ensure the agent configuration is reliable and ready for review.

An AI Skill is a structured definition file for reusable workflows, whereas Rules, Hooks, and SubAgents serve different agent configuration purposes. Skill design guidance helps determine the correct component type based on your specific configuration guidelines and operational requirements.

The best way to structure AI skill metadata is through a template-based generation process that organizes operational steps, completion criteria, and constraints. This ensures the agent configuration is validated and properly structured for review before deployment.

Yes, you can use a template to design AI workflows in Cursor. Template-based skill generation collects your requirements and configuration guidelines to produce structured definition files, reducing the complexity of designing new agent workflows for your environment.

You should not use a structured template for AI skill creation when your request does not fit the criteria of a Skill, Rule, Hook, or SubAgent. If the workflow lacks clear operational steps or completion criteria, it may not require a formal definition file.