Validate agent skills against the agentskills specification with skills-ref.
npx skills add https://github.com/bmorphism/boxxy --skill skill-bmorphism
Please help me install this Agent Skill. Skill: skill Source: https://github.com/bmorphism/boxxy/tree/main/skills/skill Command: npx skills add https://github.com/bmorphism/boxxy --skill skill-bmorphism
This Skill provides a formal specification and practical guide for creating, validating, and using agent skills. It helps teams adopt a consistent structure for SKILL.md, resource directories, and progressive disclosure.
Run validation on this skill folder with skills-ref validate, then inspect its metadata with skills-ref read-properties, and finally load its instructions when needed with skills-ref to-prompt.
High-intent search queries and answers about installing and using this skill.
Progressive disclosure for agent skills is a three-level model mapping metadata, instructions, and resources. Level 1 provides frontmatter, Level 2 contains body instructions, and Level 3 holds resource directories like scripts and references.
You validate agent skill frontmatter by running skills-ref validate on the skill folder. This enforces required fields like name and description, verifies allowed frontmatter structure, and checks the directory layout for SKILL.md compliance.
Required frontmatter fields for creating an agent skill are name and description. The specification enforces this frontmatter structure to ensure agents can properly load and reason about skill metadata during execution.
Yes, you can use skills-ref read-properties to inspect agent skill metadata before loading instructions. This follows the progressive disclosure model, allowing you to validate frontmatter first, then load instructions with skills-ref to-prompt when needed.
The correct directory layout for implementing agent skills includes a SKILL.md file at the root, alongside optional scripts/, references/, and assets/ directories. This structure supports the three-level progressive disclosure model for skill resources.
Boxxy rendering works with agent skills that pass skills-ref validation. The skill specification maps to boxxy-compatible tooling, enabling proper rendering of validated frontmatter, directory structures, and progressive disclosure content.
tvishwanadhaDefine and validate SKILL.md frontmatter conventions for reliable skill discovery.
pradyumnasagarGuide authors through creating and validating ORS SKILL.md files.
tenequmCreate Claude Skills with SKILL.md templates and YAML frontmatter.
hyperbolic-cParse YAML frontmatter from Markdown requirement files and validate required fields.
