Apply Anthropic Skill Best Practices

This is the meta-skill for writing better skills. It codifies Anthropic's official best practices into a reference you can invoke while building complex skills. The key insight here is progressive disclosure: only metadata loads upfront, so you want a laser-focused description for discovery, then structure SKILL.md as a table of contents that points to detailed files as needed. Keep the main file under 500 lines, test across models (Haiku needs more guidance, Opus needs less), and always write descriptions in third person or Claude gets confused during discovery. Use it when you're building something non-trivial and want to avoid the common mistakes that kill skill performance at scale.

skills.sh installs: 1.2K
GitHub stars: 1.8K
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add neolabhq/context-engineering-kit --skill apply-anthropic-skill-best-practices --agent claude-code

Installs into .claude/skills of the current project.

Files

Anthropic's official skill authoring best practices

Apply Anthropic's official skill authoring best practices to your skill.

Good Skills are concise, well-structured, and tested with real usage. This guide provides practical authoring decisions to help you write Skills that Claude can discover and use effectively.

Core principles

Skill Metadata

Not every token in your Skill has an immediate cost. At startup, only the metadata (name and description) from all Skills is pre-loaded. Claude reads SKILL.md only when the Skill becomes relevant, and reads additional files only as needed. However, being concise in SKILL.md still matters: once Claude loads it, every token competes with conversation history and other context.