skills
A curated collection of coding conventions and patterns for Python, Django, and web development designed to guide Claude and other AI agents toward consistent, opinionated implementations. Rather than letting AI models make arbitrary architectural decisions, this project codifies best practices for layered app structure, type hints, testing patterns, and framework-specific idioms that developers can symlink into their agent configurations.
Skills is a curated collection of coding conventions and patterns for Python, Django, and web development designed to guide Claude and other AI agents toward consistent, opinionated implementations. The project codifies best practices into structured skill files that can be symlinked into agent configurations to prevent arbitrary architectural decisions.
- ✓Novel three-tier progressive disclosure model (frontmatter, body, references) optimizes token usage while providing comprehensive guidance
- ✓Addresses a real problem of AI models making inconsistent architectural choices by encoding opinionated decisions like layered app structure and specific naming conventions
- ✓Clean installation system with symlink management and force-overwrite options makes it easy to integrate into existing AI workflows
- →Add example outputs or before/after comparisons showing how the skills actually influence AI-generated code to demonstrate concrete value
- →Include test coverage for the installation scripts and validation of the skill file formats to ensure reliability