10 Cursor Features Most Developers Don’t Know About

Plus: How to Set Up Cursor for Your First AI-Assisted Project

Cursor is one of the most powerful AI code editors out there — but most developers are only using about 20% of what it can do. Whether you’re brand new to it or you’ve been coding with it for months, there’s a good chance you’re missing some seriously useful features.

In this post, we’ll cover 10 hidden or underused Cursor features that can level up your workflow — and then walk you through setting it up for your very first AI-assisted project.


10 Cursor Features Most Developers Don’t Know About


1. Codebase-Wide Chat (@ Your Entire Project)

Most people use Cursor’s chat to ask about a single file. But you can actually ask questions about your entire codebase.

Just type @codebase in the chat and ask something like:

“Where is the user authentication logic handled?”

Cursor will search through all your files and give you a smart, context-aware answer. Great for onboarding to a new project or hunting down where something lives.


2. Multi-File Edits with Composer

The Composer feature (Ctrl/Cmd + I) lets you describe a change and have Cursor apply it across multiple files at once.

For example:

“Add error handling to all API calls in the /services folder”

Cursor will show you a diff for each file before making changes. You review, approve, and done. No more opening 10 files manually.


3. Custom AI Rules with .cursorrules

You can create a .cursorrules file in your project root to give Cursor a “personality” for your codebase.

Inside, write things like:

  • “Always use TypeScript strict mode”
  • “Prefer functional components over class components”
  • “Follow our company’s naming conventions”

Cursor reads this file automatically. It’s like giving the AI a style guide — and it actually listens.


4. Inline Code Generation with Ctrl + K

Press Ctrl + K (or Cmd + K on Mac) anywhere in your file to open an inline prompt box.

You can write things like:

“Write a function that validates email addresses”

The code appears right where your cursor is — no copy-pasting from chat required. It’s the fastest way to generate snippets mid-coding.


5. Asking Questions About Selected Code

Highlight any block of code, then press Ctrl + Shift + L to send it directly to the chat with context.

Ask things like:

“Can you explain what this does?” “Refactor this to be more readable” “Find any bugs in this function”

It’s way faster than manually copy-pasting into the chat window.


6. Referencing Docs with @docs

You can tell Cursor to reference official documentation when answering questions.

Type @docs in chat and point it to a URL — like the React docs or your own internal docs. Cursor will use that as context when generating code or explanations.

This is a game-changer when working with libraries you don’t fully know yet.


7. Git-Aware Suggestions

Cursor can see your git history and diffs. You can ask questions like:

“What changed in the last commit?” “Why was this function modified last week?”

This helps you understand changes without digging through git logs manually — especially useful in team projects.


8. Terminal AI Assistance

Cursor has a built-in terminal — and the AI can help there too.

If you’re not sure what command to run, just ask in the chat:

“How do I run this project locally?”

Cursor will suggest the exact commands based on your project files. It reads your package.json, Makefile, or requirements.txt to give accurate suggestions.


9. Privacy Mode (For Sensitive Codebases)

If you’re working on confidential code, you can turn on Privacy Mode in Cursor’s settings.

When enabled, your code is not sent to any external servers for training. This is huge for freelancers working with client code, or developers at companies with strict data policies.

Most people don’t realize this option exists.


10. Notepads (Reusable Context Blocks)

Notepads let you save chunks of context or instructions that you can pull into any conversation with @notepad.

For example, you could save your app’s architecture overview, your team’s coding standards, or a detailed bug description — and reference them anytime without rewriting.

Think of it as a clipboard that the AI can actually read and understand.


Setting Up Cursor for Your First AI-Assisted Project

If you haven’t used Cursor before, here’s how to go from zero to your first AI-powered coding session in under 10 minutes.


Step 1: Download and Install Cursor

Head to cursor.com and download the version for your OS (Windows, Mac, or Linux).

Installation is straightforward — just follow the on-screen steps. It looks and feels very similar to VS Code, so if you’ve used that before, you’ll feel right at home.


Step 2: Import Your VS Code Settings (Optional but Recommended)

On first launch, Cursor will ask if you want to import your VS Code extensions, themes, and keybindings.

Say yes. It takes about 30 seconds and saves you a ton of setup time. All your familiar tools will be ready to go.


Step 3: Choose Your AI Model

Go to Settings → Models and pick which AI model you want Cursor to use.

Options typically include Claude (by Anthropic) and GPT-4. For most coding tasks, either works well. Claude tends to be better at following complex instructions; GPT-4 is strong at general code generation.

You can switch models anytime, so don’t overthink this.


Step 4: Open or Create Your Project

Open a folder (File → Open Folder) or create a new one. This is your project workspace.

For your first AI-assisted project, start with something small — like a to-do app, a simple API, or a landing page. You want to learn the workflow before using it on something critical.


Step 5: Create a .cursorrules File

In your project root, create a file called .cursorrules.

Write 3–5 simple rules that describe your project. For example:

- This is a React + TypeScript project
- Use Tailwind CSS for styling
- Keep components small and focused
- Always add PropTypes or TypeScript types

This small step makes a huge difference in the quality of AI suggestions you get.


Step 6: Use Chat to Understand Your Codebase

Open the chat panel (Ctrl/Cmd + L) and start asking questions:

“Explain the folder structure of this project” “What does the App.tsx file do?” “How should I add a new page to this project?”

Don’t jump straight to generating code. First, make sure Cursor understands your project — and that you understand what it’s about to do.


Step 7: Generate Your First Feature with Composer

Press Ctrl/Cmd + I to open Composer. Type in a feature request:

“Create a login form component with email and password fields, basic validation, and a submit button”

Cursor will generate the code and show you exactly what it wants to add or change. Review each diff, approve what looks good, and reject anything that doesn’t fit.

Pro tip: Always review the output before accepting. AI is fast, but you’re the one responsible for the code.


Step 8: Use Inline Edits for Small Changes

For quick, small changes, use Ctrl + K instead of Composer.

Click where you want to add code, press Ctrl + K, and type a short prompt. It’s faster for things like:

  • Adding a comment
  • Writing a helper function
  • Fixing a specific line

Step 9: Ask “Why” Not Just “What”

One thing beginners miss: use Cursor to learn, not just to produce.

After generating code, ask:

“Explain what this code does line by line” “What are the tradeoffs of this approach?” “Is there a better way to do this?”

This turns Cursor into a tutor, not just a code machine — and you’ll become a better developer in the process.


Step 10: Commit Often

AI-assisted coding can move fast. That’s great — but it also means you can end up with a lot of changes quickly.

Commit your code frequently to git so you can always roll back if something goes wrong. Treat every Composer session like a mini-feature branch.


Final Thoughts

Cursor isn’t just a smarter autocomplete tool — it’s a complete shift in how you can write, understand, and maintain code. The features most people ignore (like .cursorrules, Composer, and @codebase) are often the most powerful ones.

Start small, experiment often, and don’t be afraid to ask the AI to explain its own suggestions. That’s where the real learning happens.


🔗 Want More Tips Like This?

If you found this useful, you’ll love what’s waiting for you at tooltechsavvy.com — a blog packed with practical guides on tools, tech, and smarter ways to build.

👉 Head over to tooltechsavvy.com and explore — you might find your next favourite tool waiting for you.

Leave a Comment

Your email address will not be published. Required fields are marked *