Skip to main content

Getting Started

Welcome to Corva's Dev Center! This guide walks you through everything you need to build and run a Front End app locally.

1. Prerequisites​

Before you begin, make sure you have the following:

  1. Access to the Corva Dev Center GitHub repository. You'll find Sample Apps there to help you get started.

  2. Familiarity with React, JavaScript, and TypeScript. You should be comfortable building and using React components.

  3. Basic Command Line Interface (CLI) knowledge for your operating system.

  4. An Integrated Development Environment (IDE) installed — e.g. Visual Studio Code or Cursor.

  5. Git installed for version control.

2. Install Node.js (v24.15.0)​

The Corva scaffolding tooling uses Node.js 24.15.0 (the current LTS release).

Direct install​

If you don't need to manage multiple Node versions, simply download and install Node 24 LTS from nodejs.org. If you already have Node 24 installed, you can skip this step.

Via NVM (optional)​

If you need to switch between Node versions for different projects, use a version manager:

  • Windows: Download and install nvm-setup.exe (or any other file format convenient for you) from nvm-windows. If you have an existing Node.js installation, uninstall it first to avoid PATH conflicts. Then run:
nvm install lts
nvm use lts
  • macOS / Linux: Install nvm, then run:
nvm install --lts
nvm use --lts

Verify​

node --version
# Expected: v24.15.0

3. Install Global Packages​

Install the Corva CLI tool and Yarn in one command:

npm i -g @corva/create-app yarn

or install these two packages separately 

npm i -g @corva/create-app
npm i -g yarn

Verify both are available:

create-corva-app --version
yarn -v

Windows / PowerShell users: PowerShell may block globally installed scripts by default. If yarn or npm is not recognized, run the following once and restart your terminal:

Set-ExecutionPolicy -Scope CurrentUser -ExecutionPolicy RemoteSigned -Force

This only affects your user account and does not compromise system security.

4. Update the Hosts File​

Corva's local dev server uses the hostname app.local.corva.ai. You need to point it to your machine by adding this line to your hosts file:

127.0.0.1      app.local.corva.ai

macOS / Linux​

Open a terminal and run:

sudo nano /etc/hosts

Add 127.0.0.1 app.local.corva.ai at the end of the file, save (Ctrl+O, Enter, Ctrl+X).

On macOS you can optionally flush the DNS cache afterwards:

sudo killall -HUP mDNSResponder

Windows​

  1. Open Notepad as Administrator (right-click > Run as administrator).
  2. Go to File > Open, navigate to C:\Windows\System32\drivers\etc, change the file-type filter to All Files, and open hosts.
  3. Add 127.0.0.1 app.local.corva.ai at the end, then save.

For more detailed instructions, see the Hosts File Update Guide.

5. Create an App​

Now the fun part — creating a Front End app in Corva's Dev Center using the App Wizard. Watch the video for a quick overview, then follow the steps below.

NOTE: In order to view videos in high resolution, please click on the Settings icon, then select the highest resolution available.

  1. Log in to Corva with your username and password.
  2. Click the Dev Center icon in the top-right corner of the browser window.
  3. Click Add New App.
  4. Select Front End and click Next.
  5. Give your app a name. This name appears in the Dev Center Store and is used in the app key.
  6. Select a segment and category, then click Next. Tagging your app (e.g. drilling, completions) helps users find it in the App Store.

6. Scaffold the App​

The Dev Center App Wizard generates the CLI command you need to scaffold your project. Watch the video and follow the steps below.

NOTE: In order to view videos in high resolution, please click on the Settings icon, then select the highest resolution available.

  1. From the App Wizard, copy the scaffolding command, e.g:
npx @corva/create-app app_name --appName "App Name" --segments "drilling" --category "analytics" --appKey "sample.app_name.ui" --appType "ui" --runtime "ui"

  1. Paste it into your terminal and press Enter. If you installed @corva/create-app globally in Step 3, the command runs immediately. Otherwise you can use npx @corva/create-app ... to run it without a global install.

  2. Select a programming language when prompted — JavaScript or TypeScript.

  3. The CLI generates a ready-to-use project folder.

  4. Finalize the app in the App Wizard by clicking Add New App.

If the CLI prints a message saying your version of @corva/create-app is outdated, update it first:

npm i -g @corva/create-app@latest

Then re-run the scaffolding command.

7. Run on Localhost​

Time to see your app in the browser. Watch the video and follow the steps below.

NOTE: In order to view videos in high resolution, please click on the Settings icon, then select the highest resolution available.

  1. In your terminal (or IDE), navigate into the root folder of your Front End app.
  2. Run yarn to install all dependencies.
  3. Run yarn start to start the dev server. A browser window will open at https://app.local.corva.ai:5000 with a live preview. Any saved code changes will appear automatically.

You're all set! Now you know how to build out Front End app templates using the App Wizard and your IDE. Next, explore the Dev Center Component Library for available UI components.

8. AI Tooling — Corva UI MCP Server (Optional)​

If you use Cursor as your IDE, you can enable the Corva UI MCP server. This gives Cursor's AI assistant direct access to the @corva/ui component library — it can look up component docs, browse available hooks and utilities, search for components by use case, and retrieve API client documentation, all without leaving the editor.

You don't need to do anything special to use the MCP Server. Simply ask your question in the chat as you normally would — the AI agent automatically detects when MCP tools are relevant to your request and invokes them behind the scenes.

What it provides​

The MCP server exposes these capabilities to the AI agent:

  • List all available @corva/ui items by category (components, hooks, utils, constants, clients, icons, etc.)
  • Search the component library by name or description
  • Get documentation for individual components, hooks, API clients, theme tokens, and constants

How to enable it​

Way 1 After running the scaffolding command, your project root should already contain a .cursor folder with an mcp.json file. This file has all the configuration needed to enable the MCP server.

Just restart Cursor and verify in Cursor Settings → Tools & MCPs (press the gear icon in the top pane, or Ctrl (Cmd on Mac) + Shift + J) that the corva-ui server is enabled.

Way 2 If you don't see the corva-ui MCP server, you can run

npx -p @corva/ui corva-ui-mcp-setup

in the terminal from your project's root folder.

This will set up the MCP server for you. Then enable it in Cursor Settings.

image

Option 2. Global install​

  1. Open Cursor Settings (gear icon, or Ctrl + Shift + J / Cmd + Shift + J), then navigate to Tools & MCPs in the sidebar (or search for "MCP").

  2. Click New MCP server. This will open the global mcp.json file.

  3. Set the server configuration as follows.

Add this if the file is empty:

{
  "mcpServers": {
    "corva-ui": {
      "command": "npx",
      "args": ["-p", "@corva/ui", "corva-ui-mcp"]
    }
  }
}

Or, if you already have other servers configured, add the corva-ui entry to your existing "mcpServers" block:

{
  "mcpServers": {
  ...
    "corva-ui": {
      "command": "npx",
      "args": ["-p", "@corva/ui", "corva-ui-mcp"]
    }
  ...
  }
}
  1. After saving, the server should appear under Installed MCP Servers with a green status indicator.

Once enabled, Cursor's AI agent can answer questions like "Which @corva/ui components handle loading states?" or "Show me the docs for the useSubscriptions hook" using live library data.

For broader AI-integration questions, reach out to the Dev Advocates team.

Appendix: Windows Environment Setup​

This section consolidates all Windows-specific details in one place. If you are on macOS or Linux, you can skip it.

Installing Node.js via NVM for Windows​

If you chose the NVM route in Step 2, here is the full walkthrough for Windows.

Why NVM?​

Installing Node.js directly can sometimes cause the Corva CLI to fail with advise to update Node.js to a specific version.

Using NVM lets you install and switch between Node versions without conflicts.

Install NVM for Windows​

  1. If you already have Node.js installed, uninstall it first via Settings > Apps > Installed apps. This prevents PATH and node.exe conflicts later.

  2. Download nvm-setup.exe from the nvm-windows releases page and run the installer.

  3. Open Command Prompt (cmd.exe). NVM commands may not work reliably in PowerShell, so cmd.exe is recommended for NVM operations.

  4. Install and activate Node LTS:

nvm install lts
nvm use lts
  1. Verify the installation:
node --version
# Expected: v24.15.0

Windows Troubleshooting​

node.exe not found after switching versions with NVM​

Likely causes:

  • A previous Node installation is still present — uninstall it and restart your terminal.
  • PATH or NVM environment variables are not set correctly — rerun the NVM installer.
  • The terminal was not restarted after installation — close and reopen cmd.exe.

Yarn command not recognized in PowerShell​

Run the execution-policy fix from Step 3:

Set-ExecutionPolicy -Scope CurrentUser -ExecutionPolicy RemoteSigned -Force

Then restart your terminal and try yarn -v again.

Useful NVM commands​

nvm list              # Show installed Node versions
nvm install latest    # Install the latest Node version
nvm install lts       # Install the current LTS version
nvm use lts           # Switch to the LTS version
nvm help              # Show all available commands