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.

Drift CLI understands natural language commands, so you can describe what you want to build using conversational language instead of memorizing complex ROS2 or Gazebo commands.

​
Natural language patterns

You can talk to Drift like you’re talking to a colleague. The AI understands the intent behind your commands and translates them into the appropriate robotics workflow.
No more memorizing complex ROS2 commands or Gazebo configurations — just describe what you want to build, and Drift handles the rest.

​
Common tasks

​
Creating workspaces and robots

Use natural language to create ROS2 workspaces and robot models: 
drift> create a ROS workspace

drift> create a three arm manipulator with camera sensor
Drift automatically handles:
  • Workspace and package generation
  • URDF/SDF file creation
  • Proper ROS2 structure and conventions

​
Launching simulations

Start your robot simulations with simple commands: 
drift> launch my robot with the custom world file
Drift takes care of:
  • Launch file generation and validation
  • Process management
  • Environment setup

​
Adding sensors and components

Modify your robot by describing what you want to add: 
drift> add a lidar sensor to my robot and rebuild
The CLI will:
  • Update your robot’s model files
  • Rebuild the workspace
  • Validate the configuration

​
Debugging issues

Ask Drift to help troubleshoot problems: 
drift> why isn't my camera publishing images?
Drift provides:
  • AI-assisted troubleshooting
  • Error analysis
  • Real-time problem solving
  • Adaptive planning for error recovery

​
Command examples

Here are more examples of what you can ask Drift to do: 
drift> create a new ROS2 package called my_robot
drift> set up a workspace in /home/user/robot_ws
drift> list all packages in my workspace

drift> create a mobile robot with differential drive
drift> add a depth camera to the front of my robot
drift> change the wheel radius to 0.1 meters

drift> start the simulation in Gazebo
drift> launch with the warehouse world
drift> spawn my robot at position x=1 y=2

drift> show me the sensor topics
drift> check if the lidar is working
drift> analyze the tf tree for errors

​
Getting help

If you’re unsure what to ask or want to see more examples, use the /help command: 
drift> /help
This displays comprehensive help information and command examples to get you started.

Build docs developers (and LLMs) love

Get started for freeTalk to us