mvt-create-skill

Creating custom MVTT skills requires a structured workflow that designs, generates, validates, and registers new skill units with compliant SKILL.md frontmatter and registry.yaml updates. This ensures your project-specific automation is standardized and discoverable in the MVTT catalog.

Registering a new skill in an MVTT repository requires updating the registry.yaml file for discovery after generating a complete SKILL.md with required sections. Validation checks detect invalid registry updates and name collisions before declaring the skill successfully registered.

Validating AI workflow creation involves running safety checks that detect name collisions, weak descriptions, missing standard sections, and misplaced templates. This validation occurs post-write, ensuring your MVTT skills are structurally compliant and fully functional before deployment.

MVTT skill generation supports optional resource planning by deciding whether scripts or references are needed during the guided design phase. This allows teams to define behavior and generate complete skill files tailored for reusable internal workflows.

Limitations when standardizing skills inside MVTT repositories include producing broken, duplicate, or inconsistent skills if the required structure is not followed. Misplaced templates, weak descriptions, and missing standard sections will fail validation checks during the post-write verification process.