
hexidev/planka-mcp-kanban
Developer toolsThe recommended way to run the Planka MCP server during development is through . This allows you to quickly launch and connect to the server from within your editor.
Overview
The recommended way to run the Planka MCP server during development is through . This allows you to quickly launch and connect to the server from within your editor. Create a .vscode/mcp.json file in your project root. Choose one of the following configurations depending on your use case: Use this if you want to run the published package without building locally. Replace all placeholders ( , , ) with your actual Planka credentials and API URL. You can also define these variables in a .env file or set them directly in your shell environment. Be sure to update these values to match your Planka instance. As of v1.0.3, the published package is fully self-contained and compatible with npx planka-mcp. This configuration should also work with any MCP-compatible editor (e.g., Cursor, Windsurf), though it’s currently only tested in VS Code. To run the server from your local build:
README
Planka MCP
Easily manage your Planka Kanban boards, cards, lists, and labels using an MCP (Model Context Protocol) server.
🚀 Quick Start
The recommended way to run the Planka MCP server during development is through VS Code integration. This allows you to quickly launch and connect to the server from within your editor.
1. Add MCP Configuration
Create a .vscode/mcp.json file in your project root. Choose one of the following configurations depending on your use case:
📦 For quick start / production (using published package):
{
"servers": {
"planka": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"planka-mcp"
],
"env": {
"PLANKA_EMAIL_OR_USERNAME": "",
"PLANKA_PASSWORD": "",
"PLANKA_API_URL": ""
}
}
}
}
Use this if you want to run the published package without building locally. Replace all placeholders (, , ``) with your actual Planka credentials and API URL.
Note: You can also define these variables in a
.envfile or set them directly in your shell environment.
✅ Example .env file:
PLANKA_EMAIL_OR_USERNAME=
PLANKA_PASSWORD=
PLANKA_API_URL=
Be sure to update these values to match your Planka instance.
Tip: As of v1.0.3, the published package is fully self-contained and compatible with
npx planka-mcp.
This configuration should also work with any MCP-compatible editor (e.g., Cursor, Windsurf), though it’s currently only tested in VS Code.
2. Build and Run the Server (Development Mode)
To run the server from your local build:
🛠 Development configuration (.vscode/mcp.json):
{
"servers": {
"planka": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"node",
"dist/index.js"
],
"env": {
"PLANKA_EMAIL_OR_USERNAME": "",
"PLANKA_PASSWORD": "",
"PLANKA_API_URL": ""
}
}
}
}
🛠 Features
- MCP server support for Planka v2
- Dynamic tool registration for boards, lists, cards, and labels
.env-based environment variable configuration- CLI +
npxsupport for zero-setup usage
📝 Getting Started (Development)
-
Install dependencies
npm install -
Set up environment variables Create a
.envfile or export them in your terminal:PLANKA_EMAIL_OR_USERNAME=admin PLANKA_PASSWORD=admin PLANKA_API_URL=http://example.com/api -
Build and run locally
npm run build node dist/index.jsOr, if using the published package:
npx planka-mcp
💡 Usage Examples
Run the MCP server using the default settings:
npx planka-mcp
You can also integrate the CLI with scripts or use it in any tool that supports MCP.
🖥️ Optional: VS Code Integration
To connect your MCP server to VS Code:
- Create a
.vscode/mcp.jsonfile. - Add the appropriate configuration (see “Quick Start” above).
- Start the MCP server from the VS Code command palette.
Install
npx -y planka-mcp