MCP server setup

Connect an AI assistant to PixelFixer through the Model Context Protocol.

What it enables

List teams and projects.
Read a task and the attached browser context.
Add comments and update status.
Create a task directly from an editor session.

Create a token

Open Settings → API Tokens, create a token with read and write scopes, and copy it immediately.

Add the MCP configuration

Add the configuration below to your editor's MCP settings. Replace the instance URL and token with the values for your environment.

MCP configuration

{
  "mcpServers": {
    "pixelfixer": {
      "command": "npx",
      "args": [
        "-y",
        "@mudravaorg/mcp-server"
      ],
      "env": {
        "PIXELFIXER_API_URL": "https://pixelfixer.mudrava.com",
        "PIXELFIXER_API_TOKEN": "pf_your_token"
      }
    }
  }
}

Restart and verify

Restart the assistant, then ask it to list your PixelFixer projects. If tools are unavailable, check the configuration location, token scopes, and instance URL.

If the tools are unavailable

1.Confirm the MCP client supports stdio servers.
2.Restart the assistant after changing the configuration.
3.Verify PIXELFIXER_API_TOKEN starts with pf_ and has read and write scopes.
4.Use the same instance URL that you use in the browser.
5.Check that the web service and database are reachable from the machine running the MCP server.

Use the support address for access and integration questions. tech@mudrava.comOpen the product overview