Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/drift-tech/drift-releases/llms.txt

Use this file to discover all available pages before exploring further.

​
Overview

Drift CLI eliminates the steep learning curve of robotics simulation development by letting you talk to your simulation like you’re talking to a colleague. No more memorizing complex ROS2 commands or Gazebo configurations — just describe what you want to build, and Drift handles the rest.
Natural language commands are powered by AI, making robotics development accessible to everyone from students to experienced engineers.

​
How it works

The conversational interface understands your intent and translates it into the appropriate ROS2 and Gazebo operations. You can:
  • Create and modify robot models
  • Launch simulations with custom configurations
  • Add sensors and components
  • Debug issues using plain English

Robot creation

Describe the robot you want to build, and Drift generates the necessary packages and configurations.

Sensor integration

Add sensors to your robots using natural commands without manually editing URDF files.

Simulation control

Launch and manage simulations by describing your desired setup.

Intelligent debugging

Ask questions about why something isn’t working, and get AI-powered analysis.

​
Example commands

Here are some real-world examples of natural language commands you can use with Drift:

​
Creating robots

drift> create a three arm manipulator with camera sensor
This command generates a complete ROS2 package with a three-armed manipulator robot model including an integrated camera sensor.

​
Launching simulations

drift> launch my robot with the custom world file
Drift will locate your robot and custom world file, then create and execute the appropriate launch configuration.

​
Modifying robots

drift> add a lidar sensor to my robot and rebuild
This updates your robot’s URDF, regenerates necessary files, and rebuilds the workspace automatically.

​
Debugging issues

drift> why isn't my camera publishing images?
Drift analyzes your system, checks topic connections, and provides intelligent troubleshooting guidance.

​
Workspace setup

drift> create a ROS workspace
Generates a properly structured ROS2 workspace with all necessary configuration files.
Start with simple commands and gradually add more details. Drift understands context from your previous commands.

​
Tips for effective prompts

Get the most out of Drift’s natural language interface with these best practices:

​
Be specific about components

Instead of: 
drift> make a robot
Try: 
drift> create a four-wheeled robot with a depth camera and lidar

​
Mention relevant files

When working with existing resources, reference them explicitly: 
drift> launch my turtlebot with the warehouse world

​
Ask follow-up questions

Drift maintains conversation context, so you can build on previous commands: 
drift> create a mobile robot
drift> add a camera to it
drift> now launch it in Gazebo

​
Use technical terms when needed

Drift understands robotics terminology: 
drift> configure the differential drive controller for my robot
drift> set the IMU publish rate to 100Hz

​
Request explanations

Don’t hesitate to ask “why” or “how”: 
drift> explain how the joint controllers work
drift> what topics is my robot publishing?
While Drift is powerful, it works best when you have a basic understanding of ROS2 and Gazebo concepts. Complex custom configurations may require some manual adjustments.

​
Getting started

To begin using Drift’s natural language interface:
  1. Launch Drift CLI: 
    drift
  2. Start with a simple command: 
    drift> /help
  3. Try creating your first robot: 
    drift> create a simple mobile robot
  4. Experiment and iterate - Drift learns from context!
Use the /help command anytime to see examples and get inspiration for what you can build.

Build docs developers (and LLMs) love

Get started for freeTalk to us