Skip to main content
Version: 2.1.x

Appxiom MCP Server Integration for Android

The Model Context Protocol (MCP) connects AI coding assistants and IDEs directly to Appxiom's real-time diagnostic and monitoring engine. By connecting your AI environment to the Appxiom MCP server, your coding agent can inspect real-time application crashes, ANRs, occurrence telemetry, Quality Score (QS) metrics, and Goal Friction Impact (GFI) analytics directly inside your editor.

MCP Server Configuration URL​

The Appxiom MCP server endpoint is:

https://app.appxiom.com/ax/app/api/mcp

This remote endpoint provides structured access to Appxiom tools and resources for AI agents.

Android Studio Integration​

Recent versions of Android Studio (featuring Gemini in Android Studio with Agent Mode) and the Gemini CLI provide native MCP client support, enabling Gemini to communicate directly with remote MCP servers.

Step-by-Step Setup Guide​

Step 1: Configure the MCP Server in Gemini​

Edit ~/.gemini/settings.json and add the Appxiom HTTP MCP server configuration:

{
  "mcpServers": {
    "appxiom": {
      "url": "https://app.appxiom.com/ax/app/api/mcp"
    }
  }
}

Step 2: Restart Gemini CLI or Android Studio​

Save ~/.gemini/settings.json and restart Gemini CLI (or restart Android Studio) to load the new server configuration.

Step 3: Authenticate with Appxiom​

Run the authentication command in your terminal or Gemini CLI:

/mcp auth appxiom

This will automatically open a browser window to complete the authentication flow and connect Gemini to your Appxiom account.

Step 4: Activate Gemini Agent Mode in Android Studio​

  1. Open the Gemini tool window in Android Studio (View > Tool Windows > Gemini).
  2. Switch Gemini to Agent Mode.
  3. Gemini will now automatically discover and use Appxiom MCP tools to inspect real-time issues, query crash telemetry, and assist with SDK instrumentation.

Step 5: Verify the Integration​

Verify the connection by asking Gemini in Android Studio:

  • "Gemini, check Appxiom for the top ANRs reported in the latest release."
  • "Analyze the telemetry for crash ID 4819 and inspect the breadcrumbs leading up to the issue."

Antigravity IDE Integration​

Antigravity IDE provides built-in support for remote MCP servers with automated authentication.

Step-by-Step Setup Guide​

  1. Open Settings: Open Settings → Antigravity IDE Settings.
  2. Go to Customizations: Navigate to Customizations.
  3. Open MCP Config: Click Open MCP config to open your mcp_config.json configuration file.
  4. Add Appxiom MCP Configuration: Add the Appxiom server configuration to the mcpServers object in mcp_config.json:
{
  "mcpServers": {
    "appxiom": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://app.appxiom.com/ax/app/api/mcp"
      ]
    }
  }
}
  1. Authenticate with Appxiom: After adding the configuration, Antigravity IDE will prompt you to authenticate. Follow the prompt to open the Appxiom authorization page in your browser.
  2. Paste Authentication Code: After completing authentication in the browser, copy the provided authorization code, paste it into the authentication prompt inside Antigravity IDE, and press Enter to finish connecting the MCP server.
note

In case of authentication failure, delete the server and connect again.

Claude Code Integration​

Follow these steps to connect Claude Code to the Appxiom MCP server:

Step-by-Step Setup Guide​

  1. Open your terminal to access the CLI.
  2. Run the add command: 
    claude mcp add --transport http appxiom https://app.appxiom.com/ax/app/api/mcp
  3. This will trigger an OAuth authentication flow to connect Claude Code to your Appxiom account.
  4. You may need to manually authenticate if it doesn't happen automatically, which can be done via /mcp.

Visual Studio Code Integration​

In VS Code (using MCP-enabled extensions such as Cline, Roo Code, or Continue), you can manually add the Appxiom MCP server:

Step-by-Step Setup Guide​

  1. Press Shift + ⌘ + P (macOS) or Ctrl + Shift + P (Windows / Linux) and search for MCP: Add Server.
  2. Select HTTP (HTTP or Server-Sent Events).
  3. Enter the configuration URL and hit Enter: 
    https://app.appxiom.com/ax/app/api/mcp
  4. Enter the name Appxiom and hit Enter.
  5. Allow the authentication flow to complete in your browser.
  6. Activate the server using MCP: List Servers, select Appxiom, and select Start Server.

Other AI IDE & Client Integrations​

Appxiom also supports any MCP-compliant coding assistant or IDE (such as Cursor, Claude Desktop, or Windsurf). Connect your AI environment using the remote Server-Sent Events (SSE) endpoint:

https://app.appxiom.com/mcp/sse

Complete Workflow Example​

  1. Connect & Authenticate: Configure Appxiom in Android Studio, Antigravity IDE, VS Code, or your preferred editor.
  2. Verify Connection: Ensure the MCP server connection is active in your AI tool window.
  3. Ask your Agent:
    • "Gemini, check Appxiom for the top ANRs reported in the latest release."
    • "Analyze the telemetry for crash ID 4819 and inspect the breadcrumbs leading up to the issue."
    • "Evaluate the Quality Score impact and goal friction for our checkout workflow."

Appxiom MCP Skills​

Equip your AI coding agents with pre-built skills for real-time issue analysis, occurrence telemetry investigation, Quality Score evaluation, and Goal Friction Impact tracking:

# Install all skills at once
npx @appxiom/mcp-skills add all

For complete setup instructions, available skills, and target directory options, see the MCP Skills Installation guide.