Skill Authoring
Skill Authoring empowers users to easily create and customize AI skills, streamlining development and tailoring automation to specific needs.
Install on your platform
We auto-selected Claude Code based on this skillβs supported platforms.
Run in terminal (recommended)
claude mcp add pproenca-skill-authoring npx -- -y @trustedskills/pproenca-skill-authoring
Or manually add to ~/.claude/settings.json
{
"mcpServers": {
"pproenca-skill-authoring": {
"command": "npx",
"args": [
"-y",
"@trustedskills/pproenca-skill-authoring"
]
}
}
}Requires Claude Code (claude CLI). Run claude --version to verify your install.
About This Skill
What it does
This skill, "Skill Authoring," provides guidance and best practices for creating and optimizing AI agent skills, specifically those designed for Claude Code and utilizing Model Context Protocol (MCP) tools. It offers a design and development guide containing 46 rules across eight categories to improve skill creation, review, and overall effectiveness. The goal is to ensure skills are reliable, easily discoverable, and efficient in their use of context tokens.
When to use it
- When creating new Claude Code skills or MCP tools.
- During the writing or reviewing of
SKILL.mdmetadata and descriptions. - To optimize skill trigger reliability and ensure activation.
- For structuring content using progressive disclosure techniques.
- When testing skill activation behavior and overall functionality.
Key capabilities
- Description Engineering: Guidance on crafting effective skill descriptions that drive activation through specific capabilities and keywords.
- Content Structure: Best practices for organizing skill content, including prioritizing critical instructions in the first 100 lines and utilizing progressive disclosure to save tokens.
- Trigger Optimization: Techniques for ensuring skills trigger reliably with real user phrases and edge cases.
- MCP Tool Design: Considerations for designing tool interfaces within agent workflows.
- Skill Metadata Design: Rules for formatting skill names, frontmatter, and directory structure.
Example prompts
While this skill is a guide for creating AI skills, rather than an active tool itself, a user might ask:
- "What are the best practices for writing a description to ensure my Claude Code skill activates correctly?"
- "How should I structure my
SKILL.mdfile to use progressive disclosure effectively?" - βWhat are the rules around naming conventions for AI agent skills?β
Tips & gotchas
- Prioritize Descriptions: Skill activation heavily relies on accurate and descriptive metadata. Vague descriptions can lead to inconsistent or failed activations.
- Front-load Important Information: Claude may truncate longer content, so place essential instructions and constraints within the first 100 lines of the
SKILL.mdfile. - Test Activation Thoroughly: Don't just test execution; ensure your skill triggers reliably with various user inputs.
Tags
TrustedSkills Verification
Unlike other registries that point to live repositories, TrustedSkills pins every skill to a verified commit hash. This protects you from malicious updates β what you install today is exactly what was reviewed and verified.
Security Audits
| Gen Agent Trust Hub | Pass |
| Socket | Pass |
| Snyk | Pass |
π Community
Passed automated security scans.