Intro

codeflow is the Solana AI dev agent, built for the terminal.

Let’s get started.

---

Prerequisites

To use codeflow, you’ll need:

1. A modern terminal emulator like:

   • WezTerm, cross-platform
   • Alacritty, cross-platform
   • Ghostty, Linux and macOS
   • Kitty, Linux and macOS

2. Solana CLI (solana --version) installed and configured

3. Anchor CLI for smart contract (program) development

4. Rust toolchain (for compiling Solana programs)

5. A wallet (Phantom, Solflare, Backpack, or CLI keypair)

6. An RPC endpoint (local validator, devnet, testnet, or mainnet)

7. API keys for the LLM providers you want to use.

---

Install

The easiest way to install codeflow is through the install script.

curl -fsSL https://codeflow.ai/install | bash

You can also install it with the following:

• Using Node.js

  npm

  npm install -g codeflow-ai

  Bun

  bun install -g codeflow-ai

  pnpm

  pnpm install -g codeflow-ai

  Yarn

  yarn global add codeflow-ai

• Using Homebrew on macOS and Linux

  brew install codeflow/tap/codeflow

• Using Paru on Arch Linux

  paru -S codeflow-bin

Windows

Right now the automatic installation methods do not work properly on Windows. However you can grab the binary from the Releases.

---

Configure

With codeflow you can connect directly to your Solana wallet, configure RPC endpoints, and choose your preferred AI provider.

We recommend starting with a wallet login (Phantom, Solflare, or Backpack) — it’s the most seamless way to integrate codeflow into your Solana dev workflow.

We also recommend signing up for Claude Pro or Max as a cost-effective provider for codeflow, but any supported provider will work.

Once you’ve signed up, run codeflow auth login and select your provider.

$ codeflow auth login

┌  Add credential
│
◆  Select provider
│  ● Anthropic (recommended)
│  ○ OpenAI
│  ○ Google
│  ○ Amazon Bedrock
│  ○ Azure
│  ○ DeepSeek
│  ○ Groq
│  ...
└

Alternatively, you can select one of the other providers. Learn more.

---

Initialize

Navigate to the project you want to work on:

cd /path/to/project

And run codeflow:

codeflow

Next, initialize codeflow for the project by running the following command:

/init

This will get codeflow to analyze your project and create an AGENTS.md file in the project root.

Tip

You should commit your project’s AGENTS.md file to Git.

This helps codeflow understand the project structure and the coding patterns used.

---

Usage

You are now ready to use codeflow to work on your project. Feel free to ask it anything!

If you are new to using an AI coding agent, here are some examples that might help.

---