| name | skill-creator |
| description | Create or update agent skills with proper SKILL.md structure, frontmatter, and best practices. Use when building new private skills for this repository and the vercel-labs/skills CLI. |
Skill Creator
Create well-structured agent skills that work with the open skills CLI and install cleanly into supported agents.
Skill Structure
Every skill requires this directory structure:
skills/
└── skill-name/
├── SKILL.md # Required - main skill definition
├── references/ # Optional - docs loaded only when needed
├── scripts/ # Optional - deterministic helper scripts
└── assets/ # Optional - templates or files used in outputs
SKILL.md Format
---
name: skill-name
description: Brief description of what the skill does and when to use it. Max 1024 chars.
---
# Skill Title
[Main instructions for the agent when this skill is activated]
Frontmatter Requirements
| Field | Required | Constraints |
|---|
name | Yes | Lowercase letters, numbers, hyphens only. Max 64 chars. |
description | Yes | What it does + when to use it. Max 1024 chars. |
metadata.internal | No | Set true only for skills hidden from normal discovery. |
Writing Effective Descriptions
The description determines when an agent activates the skill. Be specific:
Good descriptions:
- "Transform changelogs and user showcases into Twitter posts. Use for social media content with casual voice."
- "Design evidence-based powerlifting programs. Use for strength training, workout planning, progressive overload."
Bad descriptions:
- "Helps with content" (too vague)
- "For writing" (won't trigger reliably)
Include:
- Specific functionality and capabilities
- Trigger terms users would mention
- When Claude should activate it
Skill Content Best Practices
Structure
# Skill Title
Brief context paragraph.
## Context
Who/what this skill represents, target audience.
## [Main Sections]
Core instructions, templates, guidelines.
## Examples
Input/output examples when helpful.
## Quality Checklist
Verification criteria for outputs.
Content Guidelines
- Be specific and actionable
- Include templates and examples
- Define clear output formats
- Add quality checklists when appropriate
- Reference supporting files only when they are needed
- Keep
SKILL.md concise; move large examples or domain references into references/
Skill Locations
Use skills/<skill-name>/ as the canonical source layout in this repo. The skills CLI installs from there into each agent's preferred directory, such as ~/.codex/skills/ for Codex or ~/.claude/skills/ for Claude Code.
Creating a New Skill
- Identify the need: What task should agents handle automatically?
- Define triggers: What words/contexts should activate it?
- Create directory:
mkdir -p skills/skill-name
- Write SKILL.md: Include frontmatter + instructions
- Add supporting files: Optional reference docs
- List with CLI: Run
npx skills add . --list
- Install locally: Run
npx skills add . --global --agent codex --skill skill-name
Example: Minimal Skill
---
name: code-review
description: Review code for bugs, security issues, and best practices. Use when asked to review or audit code.
---
# Code Review Skill
Review code systematically for:
## Checklist
- [ ] Logic errors and edge cases
- [ ] Security vulnerabilities (injection, auth, etc.)
- [ ] Performance concerns
- [ ] Code style and readability
- [ ] Error handling
## Output Format
Provide findings as:
1. **Critical**: Must fix before merge
2. **Important**: Should fix soon
3. **Suggestions**: Nice to have improvements
Common Patterns
Voice/Persona Skills
Define character, tone, writing style for consistent outputs.
Task-Specific Skills
Step-by-step instructions for specific workflows.
Template Skills
Output formats, document structures, boilerplate.
Analysis Skills
Checklists, criteria, evaluation frameworks.