MagicPod MCP Server is a module that enables AI agents (Cline, Cursor, Claude, etc.) to operate various MagicPod functions. By implementing this, you can automate MagicPod operations by giving instructions to AI agents in natural language.
As of April 2025, there are mainly three capabilities:
- Execute tests via MagicPod Web API
- Retrieve test execution information (statistics, etc.) via MagicPod Web API
- Suggest solutions for usage, errors, and troubleshooting by referencing MagicPod help pages
- Create and edit test cases in natural language via MagicPod Autopilot
Limitations: Test creation, editing, and execution are only supported in cloud environments. Local PC environments are not supported.
Future expansion plans: We plan to expand functionality in the future, including enhanced automatic repair. Reference: Press Release(Japanese)
Table of contents
Setup Instructions
Here we explain how to integrate MagicPod MCP Server as a tool in Claude Desktop and Cursor. The setup procedure is the same for both MacOS and Windows.
Claude Desktop
1. Install Claude Desktop
Download Claude Desktop from here and install it on your PC.
2. Install npx
npx is a tool that comes standard with Node.js v8.2.0 and later. MCP server requires npx.
If running this command in your command-line tool displays a version number, npx is already installed, so proceed to the next step:
npx -v
If npx is not installed, install the latest version of Node.js from here.
3. Set up MagicPod MCP Server
Open (or create if it doesn't exist) the file claude_desktop_config.json and write the following:
MacOS path：~/Library/Application Support/Claude/claude_desktop_config.json
Windows path：%APPDATA%\Claude\claude_desktop_config.json
Replace "YOUR-API-TOKEN" with your API token copied from the Integrations page.
For MacOS:
{
"mcpServers": {
"magicpod-mcp-server": {
"command": "npx",
"args": ["-y", "magicpod-mcp-server", "--api-token=YOUR-API-TOKEN"]
}
}
}For Windows:
{
"mcpServers": {
"magicpod-mcp-server": {
"command": "cmd",
"args": ["/c", "npx", "-y", "magicpod-mcp-server", "--api-token=YOUR-API-TOKEN"]
}
}
}
After that, restart Claude Desktop. If you can confirm the number of installed MCP tools as shown in the image, the setup is successful.
Cursor
1. Install Cursor
Download Cursor from here and install it on your PC.
2. Install npx
npx is a command-line tool that comes standard with Node.js v8.2.0 and later. The MCP server requires npx.
If running this command in your command-line tool displays a version number, it's already installed, so proceed to the next step.
npx -v
If npx is not installed, install the latest version of Node.js from here.
3. Set up MagicPod MCP Server
Open Cursor Settings. (On Windows, you can open it from the gear icon)
Click "MCP" > "Add new global MCP server"
Write the following in the mcp.json file and save:
Replace "YOUR-API-TOKEN" with your API token copied from the Integrations page.
For MacOS:
{
"mcpServers": {
"magicpod-mcp-server": {
"command": "npx",
"args": ["-y", "magicpod-mcp-server", "--api-token=YOUR-API-TOKEN"]
}
}
}For Windows:
{
"mcpServers": {
"magicpod-mcp-server": {
"command": "cmd",
"args": ["/c", "npx", "-y", "magicpod-mcp-server", "--api-token=YOUR-API-TOKEN"]
}
}
}
If magicpod-mcp-server appears in the MCP Servers settings screen as shown in the image, the setup is successful. You can now automate MagicPod operations through the MagicPod MCP server when asking questions in the AI chat area.
Visual Studio Code
1. Install Visual Studio Code
Download Visual Studio Code (VS Code) from here and install it on your PC.
2. Install npx
npx is a command-line tool that comes standard with Node.js v8.2.0 and later. The MCP server requires npx.
If running this command in your command-line tool displays a version number, it's already installed, so proceed to the next step.
npx -v
If npx is not installed, install the latest version of Node.js from here.
3. Set up MagicPod MCP Server
Open (or create if it doesn't exist) the file mcp.json and write the following:
MacOS path：~/Library/Application Support/Code/User/mcp.json
Windows path：%APPDATA%\Code/User/mcp.json
Replace "YOUR-API-TOKEN" with your API token copied from the Integrations page.
For MacOS:
{
"servers": {
"magicpod-mcp-server": {
"command": "npx",
"args": ["-y", "magicpod-mcp-server@latest", "--api-token=YOUR-API-TOKEN"],
"type": "studio"
}
},
"inputs": []
}For Windows:
{
"mcpServers": {
"magicpod-mcp-server": {
"command": "cmd",
"args": ["/c", "npx", "-y", "magicpod-mcp-server@latest", "--api-token=YOUR-API-TOKEN"]
}
}
}Next, open Profiles in VS Code (you can open it from the gear icon in the bottom-left corner).
Click MCP Servers.
If you can confirm the number of installed MCP tools as shown in the screenshot using the mcp.json you created, the setup is successful.
Click Start to connect to the MCP server.
Prompt Examples
Usage confirmation & troubleshooting
Please explain how to handle click command failures in MagicPod.
Please explain how to exclude specific areas from the visual diff check in MagicPod.
Run tests
In MagicPod, please re-run only the failed test cases from the last batch execution in the "XX" organization's "XX" project. When re-running, explicitly specify the failed test case numbers and skip the successful cases.
In MagicPod, please execute only the file operation-related test cases in the "XX" organization's "XX" project by specifying their numbers. Use Chrome as the browser.
Investigate test run results
Get a list of failed test result URLs:
Please provide the URL list of failed test results for test case "XX" in the "XX" organization's "XX" project in MagicPod.
Investigate test failure reasons based on GitHub commit logs (requires GitHub MCP server installation):
Please investigate the following test failure: https://app.magicpod.com/OrganziationName/ProjectName/batch-run/XXX/pattern-index/X/XXX/ Please refer to this test's execution log and identify the code changes that likely caused the failure from the commit logs in the <GitHub Organization>/<GitHub Project> repository on MM/DD.
Test Execution Result Analysis:
In MagicPod's "XX" organization's "XX" project, please organize and output the results by test settings in CSV format based on batch execution results from the past 7 days.
Create a test case
Please create the following test using Autopilot: Organization name: XX Project name: XX Test settings number: X Please create steps to access <URL>, enter the following information, and complete the member registration. Email address: test@example.com Password: password
Edit a test case
Please edit the following test using Autopilot: Organization name: XX Project name: XX Test case number: X Please change the email address for registration to magicpod@example.com.
Identifying Unstable Locators
The following page introduces a method for identifying unstable locators within test cases using an MCP server.
About "Utilization of stable locators"
Other
Please provide the health score of MagicPod's "XX" organization's "XX" project.
In MagicPod's "XX" organization's "XX" project, I want to delete .app files that meet all these conditions: - The last upload date was more than 3 months ago - A newer version with the same filename already exists First, please extract a list of files to be deleted. After I confirm the content and reply with "approved", please execute the deletion process.
Troubleshooting
Error Displayed During Response Generation
For example, when using the free version of Claude, you may see an error like "Claude's response was interrupted" during response generation. This is due to character limits in the free version and can be avoided by using the paid Claude Pro version. Similar issues may occur with other AI agents, so consider subscribing to paid plans.
Security
MagicPod MCP Server is currently a very simple module that connects AI agents with MagicPod services (MagicPod Web API and Help Center). The repository is publicly available on GitHub.
- Your information will not be used for machine learning.
- While the Web API token needs to be written in the MCP server's config file, it is designed not to pass this information to AI agents.
- We plan to expand functionality in the future while maintaining a security-conscious configuration.