README.md•6.93 kB
# Installation guide
## Configure Visual Studio Code Copilot
This tutorial will guide you through the process of configuring Visual Studio Code Copilot, an AI-powered code completion tool for developers. Copilot integrates seamlessly with VS Code to provide intelligent code suggestions, accelerate development, and improve productivity across multiple programming languages.
### Prerequisites
Before you begin, ensure you have the following:
- ROS 2 installed on your system (humble or later). [Get ROS 2](https://docs.ros.org/en/humble/Installation.html)
- Docker installed and running. [Get Docker](https://docs.docker.com/get-docker/)
- Visual Studio Code installed. [Get Visual Studio Code](https://code.visualstudio.com/Download)
### Step 1: Install ROS2 MCP
1. Click on the `tools` icon
2. Click on the `MCP` icon
3. Choose type of the MCP server as `Docker Image`
4. Enter the name of the image `mcp/ros2`
## Configure Claude Desktop
This tutorial will guide you through the process of configuring Claude Desktop, a modern AI assistant available on Windows and macOS. Claude provides natural language interaction, context-aware reasoning, and productivity tools designed to streamline your workflow and enhance creativity directly from your desktop.
### Prerequisites
Before you begin, ensure you have the following:
- ROS 2 installed on your system (humble or later). [Get ROS 2](https://docs.ros.org/en/humble/Installation.html)
- Docker installed and running. [Get Docker](https://docs.docker.com/get-docker/)
- Claude Desktop installed. [Get Claude Desktop](https://claude.ai/download)
### Step 1: Install ROS2 MCP
1. Open `Claude` and navigate to the settings.
2. In the settings menu, select `Developer from the sidebar.
3. Click on the `Edit config` in `local MCP servers`.
4. Paste it into config:
```json
{
"ros2_mcp": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"wisevision/ros2_mcp:<humble/jazzy>"
],
"env": {},
"working_directory": null,
"start_on_launch": true
}
}
```
To use custom messages [create folder](#add-custom-messages) and paste it into config:
```json
{
"ros2_mcp": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-v", "~/mcp_custom_messages:/app/custom_msgs",
"wisevision/ros2_mcp:<humble/jazzy>"
],
"env": {},
"working_directory": null,
"start_on_launch": true
}
}
```
To use custom prompts:
```json
{
"mcp_server_ros_2": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e", "MCP_CUSTOM_PROMPTS=true",
"wisevision/mcp_server_ros_2:<humble/jazzy>"
],
"env": {},
"working_directory": null,
"start_on_launch": true
}
}
```
To use custom prompts from local folder:
```json
{
"mcp_server_ros_2": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e", "MCP_CUSTOM_PROMPTS=true",
"-e", "MCP_PROMPTS_LOCAL=true",
"-v", "<path_to_folder>/ros2_mcp_prompts:/app/ros2_mcp_prompts",
"wisevision/mcp_server_ros_2:<humble/jazzy>"
],
"env": {},
"working_directory": null,
"start_on_launch": true
}
}
```
### Step 2: Save and Enjoy
After saving you should see indicator that the MCP server is running.
## Configure WARP
This tutorial will guide you through the process of configuring WARP, a modern terminal emulator for Windows, macOS, and Linux. WARP offers a sleek interface and powerful features that enhance your command-line experience including MCP support.
### Prerequisites
Before you begin, ensure you have the following:
- ROS 2 installed on your system (humble or later). [Get ROS 2](https://docs.ros.org/en/humble/Installation.html)
- Docker installed and running. [Get Docker](https://docs.docker.com/get-docker/)
- WARP installed. [Get WARP](https://www.warp.dev/)
### Step 1: Install ROS2 MCP
Install by cloning the repository and building the package.
```bash
git clone https://github.com/wise-vision/ros2_mcp.git
cd ros2_mcp
docker build -t wisevision/ros2_mcp .
```
### Step 2: Configure WARP
1. Open WARP and navigate to the settings by clicking on the gear icon in the bottom left corner.
2. In the settings menu, select AI from the sidebar.
3. Scroll down to the MCP Servers section.
4. Click on Add to create a new server configuration.
### Step 3: Add MCP Server
```json
{
"ros2_mcp": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"wisevision/ros2_mcp:<humble/jazzy>"
],
"env": {},
"working_directory": null,
"start_on_launch": true
}
}
```
To use custom messages [create folder](#add-custom-messages) and paste it into config:
```json
{
"ros2_mcp": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-v", "~/mcp_custom_messages:/app/custom_msgs",
"wisevision/ros2_mcp:<humble/jazzy>"
],
"env": {},
"working_directory": null,
"start_on_launch": true
}
}
```
To use custom prompts:
```json
{
"mcp_server_ros_2": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e", "MCP_CUSTOM_PROMPTS=true",
"wisevision/mcp_server_ros_2:<humble/jazzy>"
],
"env": {},
"working_directory": null,
"start_on_launch": true
}
}
```
To use custom prompts from local folder:
```json
{
"mcp_server_ros_2": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e", "MCP_CUSTOM_PROMPTS=true",
"-e", "MCP_PROMPTS_LOCAL=true",
"-v", "<path_to_folder>/ros2_mcp_prompts:/app/ros2_mcp_prompts",
"wisevision/mcp_server_ros_2:<humble/jazzy>"
],
"env": {},
"working_directory": null,
"start_on_launch": true
}
}
```
### Step 4: Save and Enjoy
After saving you should see indicator that the MCP server is running. You can now test the setup by using the AI features in WARP.
# Build docker image locally:
```bash
git clone https://github.com/wise-vision/ros2_mcp.git
cd ros2_mcp
docker build -t ros2_mcp:<humble/jazzy> --build-arg ROS_DISTRO=<humble/jazzy> .
```
---
### Thanks for using our MCP server for ROS 2! Good luck with your projects!
---
## Add custom messages
### Create folder for custom messages
```bash
mkdir -p ~/mcp_custom_messages/src
```
### Add messages packages to `src` in `mcp_custom_messages`
```bash
cd ~/mcp_custom_messages/src
git clone https://github.com/ros/ros_tutorials.git #or your custom message pack
```
### Build
```bash
cd ~/mcp_custom_messages
colcon build
```