Cookbook Audit

About

This skill provides a professional framework for reviewing and scoring Anthropic Cookbook notebooks, ensuring they meet rigorous standards for technical accuracy, narrative quality, and security. It automates technical checks like secret detection and notebook validation while guiding the user through a qualitative audit based on the Diataxis framework. By converting notebooks to streamlined markdown for review and providing structured reports, it helps developers create educational content that is actionable, secure, and educationally sound.

Key Features

• Standardized 20-point scoring rubric across narrative and code quality
• Validation of Anthropic-specific model naming and API patterns
• Automated secret detection for hardcoded API keys and credentials
• Structured report generation with prioritized improvement recommendations
• Markdown conversion of notebooks for optimized context-aware review
• 0 GitHub stars

Use Cases

• Updating legacy AI tutorials to modern API standards and model names
• Providing structured feedback to community contributors on code examples
• Reviewing technical notebooks before publication to a public repository