Open Source · Trilingual Learning Log
claude-study
Translates Claude Code into the no-code vocabulary most knowledge workers already know.
22 chapters + cheatsheet · 6 stages · KR / EN / JP parity · 2026–
What it is & why
Claude Code is documented for developers. Most knowledge workers approach it from a no-code mental model — Zapier, Make, Cursor, Notion AI, ChatGPT Memory. The fastest on-ramp is to translate every concept into the vocabulary they already use, not the vocabulary developer docs assume.
claude-study is that on-ramp, written from the seat of someone learning Claude Code from no-code. Every chapter follows the same shape: concept → no-code comparison → how it actually works → practical scenario. Korean original, with complete English and Japanese translations.
The translation move
Every chapter starts from a tool the reader already knows. The goal is to let Claude Code feel like an extension of their existing toolkit, not a foreign country.
| Tool you know | Maps to |
|---|---|
| Zapier / Make | Skills + MCP for automation |
| Notion AI | CLAUDE.md for project rules |
| Cursor | Terminal access to the whole filesystem |
| ChatGPT Memory | Auto Memory + CLAUDE.md |
| Chrome extensions | Plugin Marketplace |
Roadmap of chapters
- Stage 1 Foundations — how Claude Code works Ch 01–03
- Stage 2 Expansion — splitting and connecting work Ch 04–06
- Stage 3 Practical — external integration and automation Ch 07–10
- Stage 4 Deep dive — production-grade decisions Ch 11–16
- Stage 5 Latest — Claude Code as it evolves Ch 17–19
- Stage 6 Internals — how it is built Ch 20–22
- Ref Terminal command cheatsheet Ch 99
Status
Public on GitHub. Markdown-only, no build step. Continues a thread that started with the IP-licensed Metacode "ChatGPT for Business" curriculum (2024) — both pieces written from the same belief that the bottleneck for AI adoption among non-developers is vocabulary, not capability.