Cursor installed and configured: your preferred model set, a project opened, and your first AI-assisted code written using Tab completion and inline editing.
What Makes Cursor Different
Cursor is a fork of VS Code — it looks identical and supports all the same extensions. The difference is how AI is integrated. Other tools add AI as a sidebar feature. Cursor makes AI the primary interaction model.
GitHub Copilot
→ In-line completion only
→ Suggests next line based on context
→ Doesn't understand full codebase
ChatGPT / Claude.ai
→ Copy-paste workflow (slow)
→ No IDE integration
→ Doesn't see your actual files
Cursor
→ Tab: multi-line completion based on intent
→ Cmd+K: inline editing with AI
→ Cmd+L: full chat with codebase context
→ Composer: write entire features across files
→ @mentions: reference specific files, docs, web pagesThe key advantage: Cursor knows your codebase. You can ask "how does user authentication work in this project?" and it reads the relevant files to answer. Copilot and ChatGPT can't do this.
Install and Configure
Download Cursor from cursor.so. Install it like any application. On first launch:
Import your VS Code settings if prompted — this brings over your extensions, themes, and keybindings so Cursor feels immediately familiar.
Open Settings (Cmd/Ctrl + ,) → Cursor Settings. Under Models, choose your preferred AI. Claude Sonnet or GPT-4o are the best choices for coding in 2025.
Enable Auto-import in settings if you use TypeScript or JavaScript — Cursor will automatically add import statements when it writes code that needs them.
Free vs Pro: Cursor's free tier gives you 2,000 AI completions/month. For serious development, the $20/month Pro plan is worth it — unlimited completions and access to better models. Try free first and upgrade when you hit the limit.
Tab Completion and Inline Editing
Tab Completion. This is the feature you'll use most. Start typing or write a comment describing what you want, and Cursor predicts your next lines. Press Tab to accept, Escape to reject.
The key difference from Copilot: Cursor's Tab completion is "next edit prediction" — it predicts what change you're about to make based on recent edits, not just what line comes next. Write one function, then start writing a similar one — Cursor fills it in based on the pattern.
Inline Editing (Cmd+K). Select any code and press Cmd+K (Ctrl+K on Windows). A small prompt appears. Type what you want: "add error handling," "refactor to use async/await," "add type hints." The AI modifies the selected code in place, and you see a diff before accepting.
"add try/except for network errors"
"convert this to a list comprehension"
"add docstring"
"make this function take a list instead of a single item"
"optimize this loop"Practice both on a real file you're working on. The goal for today: understand when Tab feels natural vs. when Cmd+K is better. Tab for continuation. Cmd+K for transformation.
What You Learned Today
- Why Cursor is different from Copilot and chat tools — codebase awareness changes everything
- How to install and configure Cursor: model selection, VS Code settings import
- Tab completion: predicts your next edit based on patterns, not just next lines
- Inline editing with Cmd+K: transform selected code with a natural language instruction
Go Further on Your Own
- Open a file from a real project and use Tab completion for 15 minutes. Note: when does it nail it, and when is it wrong? What patterns do you notice?
- Use Cmd+K to add docstrings to 5 functions in a Python file. How accurate are the generated docstrings?
- Compare Cursor's Tab completion speed to your normal typing speed on a task you've done before. How much faster is it?
Nice work. Keep going.
Day 2 is ready when you are.
Continue to Day 2Want live instruction and hands-on projects? Join the AI bootcamp — 3 days, 5 cities.