The Rundown: In this guide you'll learn how to use Cursor 2.0's multi-agent feature to build web applications faster by having different AI agents work on separate parts of your project simultaneously.

Who is this useful for:

• Solo developers who want to ship features faster
• Founders prototyping MVPs quickly to validate ideas
• Teams coordinating work on different app components
• Anyone tired of building apps one feature at a time

STEP 1: Create your project and open it in Cursor

Create a new folder on your computer and name it after your project. For example, we'll build a "todo-list-app."

Open the Cursor application and click the "Open Project" button and select the folder you just created to open it in Cursor IDE. Alternatively, you can simply drop the project folder to open it in Cursor IDE.

STEP 2: Generate a development plan with an agent

Click the "Agents" button in the top-left corner of Cursor IDE. Click "New Agent" to start a conversation.

Prompt the agent to create a development plan:

"Create a plan to build a basic To-Do list app using Next.js and output it as a Plan.md file. Consider that we have two dev agents who can work on different non-dependent aspects in parallel. Include tasks for each agent and final integration steps, plus a README for running the application."

Cursor asks clarification questions about your tech stack. Reply, based on your preferences. The agent generates a Plan.md file with the complete development strategy divided between two agents.

STEP 3: Build in parallel using multiple agents

Review the Plan.md file and make any changes you need, then accept it.

Click the "New Agent" button to start the first developer agent and prompt it to work on the specific tasks:

"Start working on the Dev 1 tasks from the Plan.md file"

Again, click the "New Agent" button to start a second developer agent. You can prompt:

"Start working on the Dev 2 tasks from the Plan.md file"

Both agents now work simultaneously on different parts of your app. Once they complete their tasks, click the "Review" button for each agent to see all changes made. Review and accept the changes from Dev Agent 1, then do the same for Dev Agent 2.

STEP 4: Integrate and run your application

Ask one of the agents to complete the final integration:

"Complete the integration steps from the Plan.md file and update the README with instructions to run the application."

Once finished, open the terminal in Cursor and run these two commands one after the other:

npm install

This will install all the necessary dependencies to run this on your system.

npm run dev

Your "To-Do list" app will start running locally at http://localhost:3000. You just built a complete application with multiple AI agents handling different pieces simultaneously.

Pro tip: Use the Voice button on the prompt bar to give instructions hands-free - just speak your prompt and end with the word "submit" to send it.