What is skills?

This repository is a collection of skills for AI coding assistants, made by a developer educator who has been teaching software concepts for over a decade. The skills are intended for developers, designers, and engineers who use AI to generate code and want to stay genuinely knowledgeable about what they are shipping. The core idea is that people who put AI-generated code into production should be able to understand and be accountable for it. The project is a response to concerns that developers using AI tools may lose depth of understanding by simply accepting code they did not write themselves. The only skill documented so far is called do-i-understand. It works as a reverse code review: rather than having another person review your code, an AI looks at the changes you are about to push and then questions you about them. It asks about your reasoning, the side effects of the changes, the assumptions baked in, and the ways things could go wrong. The point is to surface any parts of the code you have not actually thought through. At the end, it produces a report you can attach to a pull request description or keep for your own records. The README suggests three uses: personal skill maintenance, creating a structured learning process for junior engineers who are permitted to say "I don't know" as part of the exercise, and as a lightweight requirement for open source contributors. Installing the skill requires running a short command in a terminal using npx. No other setup details are provided in the README.