ChenChenyaqi/learn-anything

What it solves

It transforms AI coding assistants into interactive tutors by providing a structured, recursive learning system. It solves the problem of fragmented learning when using AI, replacing simple Q&A with a systematic approach involving knowledge maps, spaced repetition, and TDD-style practice.

How it works

The system uses a CLI tool to generate specific skill and command files (using an adapter pattern for formats like YAML, TOML, and Markdown) for over 30 different AI coding tools (e.g., Claude Code, Cursor, Gemini). These files enable six specialized slash commands within the AI assistant to manage learning paths, recursive explanations, coding exercises, and quizzes. Learning progress is stored in a local .learn directory, which can be visualized via a built-in web dashboard.

Who it’s for

Developers and learners who use AI coding assistants and want a more pedagogical approach to mastering technical topics through structured guidance and active practice.

Highlights

• Recursive Learning: Allows users to dive deeper into topics based on their curiosity.
• TDD-style Practice: Provides structured coding exercises with feedback.
• Multi-Tool Support: Compatible with 30+ AI assistants including Claude Code, Cursor, and GitHub Copilot.
• Visual Dashboard: A zero-config web interface to browse knowledge maps, session notes, and exercise results.
• Spaced Repetition: Smart review system to surface weak spots.
• Context7 Integration: Optional documentation verification to improve teaching accuracy by cross-referencing official sources.