Quick Start

This page walks through the minimum steps to get a robot project running with the Raccoon CLI.

How the five steps connect

sequenceDiagram
    participant L as Laptop
    participant Pi as Wombat Pi

    L->>L: raccoon-setup script (new project, uv sync)
    L->>Pi: raccoon connect 192.168.4.1 (SSH key setup)
    Pi-->>L: API token saved
    L->>L: raccoon run (codegen locally)
    L->>Pi: push files via rsync/SFTP
    Pi->>Pi: execute src/main.py
    Pi-->>L: stream stdout live
    Pi-->>L: pull calibration/logs after run

After raccoon connect succeeds once, every subsequent raccoon run is fully automated — no password prompts, no manual file copying.

Prerequisites

  • Python 3.13 or newer
  • git on your PATH (needed for project creation and checkpoints)
  • Your laptop on the same network as the Wombat robot

Install

pip install raccoon-cli

Create a project

A project is a git repository made by the setup script of raccoon-example. The CLI itself does not create projects.

# Linux / macOS
curl -LsSf https://raw.githubusercontent.com/htl-stp-ecer/raccoon-example/main/raccoon-setup.sh | bash

# Windows (PowerShell)
irm https://raw.githubusercontent.com/htl-stp-ecer/raccoon-example/main/raccoon-setup.ps1 | iex

The script installs git and uv if they are missing, asks for a project name and a drivetrain, clones the matching template into a new folder, gives it a fresh UUID and git history, and runs uv sync. Then edit raccoon.project.yml for your hardware; raccoon is installed in the project’s own environment and is used from inside the project folder.

Connect to the robot

raccoon connect 192.168.4.1

Replace the IP with the address shown in BotUI. raccoon will check connectivity, set up SSH key authentication (one-time password prompt), and save the address to both the project config and ~/.raccoon/config.yml.

Default Wombat credentials: user pi, password raspberrypi.

Run the project

cd MyRobot
raccoon run

raccoon will:

  1. Validate the project config
  2. Generate defs.py, defs.pyi, and robot.py from your YAML definitions (runs locally on the laptop)
  3. Push the project files to the Pi
  4. Execute the project on the Pi

Verify your environment

raccoon doctor

This checks that all required tools (ssh, git, paramiko, black) are present, shows the current connection state, and compares installed package versions against the bundle manifest.

Next steps

Written by Florian Schwanzer
create →