Installation

Set up AgentFlow on your machine. The app runs as a native desktop application built with Tauri, requiring both Rust and Node.js toolchains.

Prerequisites

Make sure the following are installed before building AgentFlow:

Rust 1.77.2+

Backend runtime and compiler

curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh

Node.js 20+

Frontend build toolchain (npm included)

nvm install --lts

Tauri CLI v2

Desktop app framework CLI

cargo install tauri-cli --version ^2.0

Claude Code CLI

Required at runtime for AI task execution

npm install -g @anthropic-ai/claude-code

Platform support

AgentFlow builds for Windows, macOS, and Linux via Tauri v2. Tauri has its own platform-specific prerequisites (e.g., webkit2gtk on Linux, Xcode CLT on macOS).

System Dependencies

Linux (Debian / Ubuntu)

sudo apt update
sudo apt install -y libwebkit2gtk-4.1-dev \
  libgtk-3-dev libjavascriptcoregtk-4.1-dev libsoup-3.0-dev \
  build-essential curl wget file patchelf \
  libssl-dev libappindicator3-dev \
  librsvg2-dev

macOS

xcode-select --install

Windows

Install Visual Studio Build Tools with the "Desktop development with C++" workload and the Windows 10/11 SDK.

Clone & Build

Clone the repository

bash

git clone https://github.com/jadessoriano/agent-flow.git
cd agent-flow

Install frontend dependencies

bash

npm install

Build the desktop app

bash

npm run tauri build

The built application will be in src-tauri/target/release/bundle/ with platform-specific installers (.deb, .dmg, .msi, etc.).

Development Mode

For local development with hot module replacement:

Start the dev server

bash

npm run tauri dev

This starts the Vite dev server on localhost:1420 with HMR and opens the Tauri desktop window pointed at it.

Frontend only

If you only want to iterate on the UI without the Tauri shell, you can run npm run dev alone. IPC calls to the Rust backend will fail, but layout and styling work fine.

Data Locations

Data	Location
SQLite database	`~/.cache/com.agentflow.app/agentflow.db`
Settings	`{APP_DATA_DIR}/settings.json`
Application logs	`~/.cache/com.agentflow.app/agentflow.log`
Agents	`{project}/.claude/agents/*.md`
Pipelines	`{project}/.claude/pipelines/*.pipeline.json`

Verify Installation

After building, launch the app and check that:

The application window opens to the project selection screen.
The Claude CLI is auto-detected (check Settings → Claude CLI Path).
Pointing to a git repository with a .claude/ directory loads any existing agents and pipelines.
Creating a new pipeline shows the visual canvas with the node palette.

Claude CLI required at runtime

Building AgentFlow does not require the Claude CLI, but running AI task nodes does. The app will warn you if the CLI cannot be found when you try to execute a pipeline.

Quickstart

Configuration