# Skill Authoring > Design, draft, and refine SKILL.md-based agent skills with clear trigger criteria, execution flow, validation checks, and reusable references. Use when creating a new skill or improving an existing one. Use when this capability is needed. - Skill: `tomevault-io/skill-authoring-3` (Agent Skill, multi-file: 2 files) - Install (CLI): `npx skillmds@latest add tomevault-io/skill-authoring-3` - Raw SKILL.md: https://api.skillmd.com/api/skills/tomevault-io/skill-authoring-3/raw - Safety review: pending (external: skill-scanner PASS, skillspector PASS) - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: AI & ML - Author: tomevault-io (https://skillmd.com/u/tomevault-io) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/tomevault-io/skill-authoring-3 --- # Skill Authoring Define skill scope, build concise instructions, and package reusable context. ## Workflow 1. Extract user intent and expected artifacts. 2. Ask only the minimum clarifying questions required to avoid rework. 3. Draft frontmatter with explicit trigger conditions. 4. Write procedural body instructions with validation and failure handling. 5. Move domain-heavy detail into `references/`. 6. Add deterministic utilities in `scripts/` when repetition exists. ## Output Requirements - Include required frontmatter fields. - Keep action steps ordered and testable. - Include at least one verification section. - Avoid redundant reference duplication. ## References Load `references/checklist.md` before finalizing output. --- > Converted and distributed by [TomeVault](https://tomevault.io/claim/lvcoi) — claim your Tome and manage your conversions.