AI Tooling Setup

You don’t need to know how to code to build with Arete. AI coding tools can write all the code for you — you just tell them what you want in plain English.

Arete works with any AI coding agent. This guide uses Cursor as the example, but the same steps apply to Claude Code, Windsurf, OpenCode, or any other agent you prefer.

What You’ll Need

A computer (Mac, Windows, or Linux)
An internet connection
About 10 minutes

That’s it. No programming experience required.

Step 1: Install Cursor

Cursor is an AI-powered code editor — a good starting point if you don’t already have a tool you prefer.

Go to cursor.sh and download the version for your OS
Open the downloaded file and follow the installation prompts
Launch Cursor and create a free account

Step 2: Create a Project Folder

Create a folder for your project and open it in your AI tool.

On Mac: Open Finder, navigate to Documents (or wherever you keep projects), right-click → New Folder, name it my-arete-app, then open it in Cursor via File → Open Folder.

On Windows: Open File Explorer, navigate to Documents, right-click → New → Folder, name it my-arete-app, then open it in Cursor via File → Open Folder.

On Linux / terminal-based tools:

mkdir ~/my-arete-app
cd ~/my-arete-app

Then open the folder in your editor, or start your agent in that directory.

Step 3: Let Your AI Set Up Arete

Paste this prompt into your AI assistant’s chat:

Read https://docs.arete.run/agent.md and follow it to set up Arete in this project, then tell me what live data is available.

In Cursor, press Cmd+L (Mac) or Ctrl+L (Windows) to open the AI chat sidebar, paste the prompt, and press Enter. Approve any commands it asks to run.

For terminal-based agents (Claude Code, OpenCode, etc.), just paste it directly and press Enter.

What happens next

Your AI will:

Install the Arete CLI — a prebuilt binary via curl -fsSL https://arete.run/install.sh | sh (or irm https://arete.run/install.ps1 | iex on Windows). No Rust, no account.
Run a4 init -y — installs the agent skills (teaching itself how to use Arete correctly) and configures the Arete MCP servers for your editor
Run a4 doctor --json — verifies everything is in place
Discover available data — a4 explore --json lists what blockchain data you can access
Tell you what it found — and wait for your next instruction

This takes about 1–2 minutes.

Step 4: Scaffold the ORE Template

Run this in your terminal to create a working ORE dashboard:

npx @usearete/a4 create my-ore-app --template react-ore
cd my-ore-app
npm install
npm run dev

Open localhost:5173 and you’ll see live ORE mining data streaming from Solana.

Step 5: Customise It With Your AI

Now open the project in Cursor and try this prompt:

I have a Arete ORE dashboard running using the react-ore template. Update the mining grid so each square is coloured as a heatmap based on the amount of SOL deployed to that square — more SOL should glow brighter. Keep the existing dark theme.

Your AI already knows how to use Arete from Step 3 — it will find the right views and update the component for you.

Troubleshooting

a4: command not found — The installer put a4 in ~/.local/bin and printed A4_BIN=<path>; your terminal captured PATH before that. Open a new terminal, or run export PATH="$HOME/.local/bin:$PATH". a4 doctor reports this as the cli.path check.

npx: command not found or skills: skipped — Skills need Node.js. Download the LTS version from nodejs.org, install it, then run a4 doctor --fix (or re-paste the setup prompt).

The AI seems confused — Make sure it completed Step 3. You can re-paste the setup prompt to re-run it.

Next Steps

Build a Dashboard

Follow our step-by-step tutorial to build a complete ORE mining dashboard.

Start tutorial →

Copy-Paste Prompts

Browse our cookbook of ready-to-use prompts for common tasks.

View prompts →