Skip to main content

MCP Server

Overview​

The CloudAware MCP (Metadata and Query Proxy) Server provides a secure interface for discovering and querying metadata and data from your CloudAware datasets stored in Google BigQuery. It enables programmatic access to object types, fields, relationships, and direct SQL query execution, helping you explore and analyze your cloud infrastructure data efficiently. Note that the server is currently under active development, and its features, APIs, and behaviors may evolve significantly in future releases. For the latest updates, refer to the release notes or contact support.

Tools Overview​

The server exposes a set of tools via a JSON-RPC interface for metadata discovery and data querying. These tools are designed to be used in a step-by-step manner:

  • search_types: Search for object types based on keywords to identify API names and table IDs.
  • search_fields: Retrieve field metadata for a specific type, including labels, types, and descriptions.
  • get_relationship_graph: Get a graph of relationships starting from a type to understand joins.
  • analyze_field: Provides statistics about the values of a specific field of a specific type.
  • execute_query: Run custom BigQuery SQL queries on your dataset.

AI agent should follow a discovery-first approach: start with type and field searches before querying data. Detailed tool descriptions, inputs, and outputs are available in the API reference (accessible via the server's endpoint).

Configuration Parameters​

The MPC server requires 2 parameters to be accessible to operate:

  • exportProject - The GCP project where the data exported from CloudAware is located.
  • sObjectsDataset - The dataset within the exportProject where all the tables are located.

Depending on the environment where the server is running, these parameters are retrieved differently:

  • Locally run servers use repo-manager and gcloud commands to obtain the values
  • Servers that run with OAuth and HTTP transport require these parameters to be provided explicitly via query parameter or HTTP header (see below)

You also have an ability to switch the exportProject and sObjectsDataset on the fly, by asking your agent explicitly:

Use export project my-export-project and dataset my-dataset. How many EC2 Instances I have?

The agent should start using these new values for all subsequent requests to the MCP server.

stdio Transport​

The MCP Server supports local operation though stdio transport.

The server is run through the repo-manager tool.

This server uses:

You have to set up both authentication profiles before using the server.

Claude Code​

Add Cloudaware MCP by running the following command:

claude mcp add --transport stdio cloudaware-mcp -- java -jar ~/.ca/repo-manager.jar mcp cloudaware

Assuming you have repo-manager installed in ~/.ca/repo-manager.jar. If not, provide the path to your repo-manager.jar file.

After adding the MCP server, start Claude Code and use /mcp command to authenticate.

Gemini CLI​

Add following to your ~/.gemini/settings.json:

{
  "mcpServers": {
    "cloudaware-mcp": {
      "command": "repo-manager",
      "args": ["mcp", "cloudaware"]
    }
  }
}

Or in case you haven't created alias for repo-manager:

{
  "mcpServers": {
    "cloudaware-mcp": {
      "command": "java",
      "args": ["-jar", "path/to/your/repo-manager.jar", "mcp", "cloudaware"]
    }
  }
}

Kilo Code​

Configuration is the same as for the Gemini CLI. The configuration file is usually located in .kilocode/mcp.json in your working directory.

OpenCode​

Add following to your opencode.json:

{
  "mcp": {
    "cloudaware-mcp": {
      "type": "local",
      "command": ["repo-manager", "mcp", "cloudaware"],
      "enabled": true
    }
  }
}

Or in case you haven't created alias for repo-manager:

{
  "mcp": {
    "cloudaware-mcp": {
      "type": "local",
      "command": ["-jar", "path/to/your/repo-manager.jar", "mcp", "cloudaware"],
      "enabled": true
    }
  }
}

HTTP Transport Running Locally​

Similarly to stdio transport, you can run the server locally with repo-manager mcp cloudaware --port 8888. This will start the server on port 8888.

LM Studio​

Edit mcp.json and add following:

{
  "mcpServers": {
    "cloudaware-mcp": {
      "url": "http://localhost:8888/mcp"
    }
  }
}

HTTP Transport With OAuth​

For tools that support integration with HTTP transport and OAuth authentication, you can use the server's endpoint URL to make requests directly.

Configuration Parameters:

  • Server URL: https://inbound.prod.cloudaware.com/mcp
  • Export Project ID: set via X-CA-ExportProject header
  • SObjects Dataset Name: set via X-CA-SObjectsDataset header

Server uses Google OAuth 2.0 for authentication. You must have required permissions to access SObjects dataset in Export Project.

Claude Code​

Add Cloudaware MCP by running the following command:

claude mcp add --transport http cloudaware-mcp "https://inbound.prod.cloudaware.com/mcp?exportProject=your-export-project-id&sObjectsDataset=your-sobjects-dataset-name"

Replace your-export-project-id and your-sobjects-dataset-name with your actual values.

After adding the MCP server, start Claude Code and use /mcp command to authenticate.

Claude.ai​

Add Cloudaware MCP by going into Settings → Connectors → Add custom connector and adding a new server with the following parameters:

  • Name: cloudaware-mcp
  • URL: https://inbound.prod.cloudaware.com/mcp?exportProject=your-export-project-id&sObjectsDataset=your-sobjects-dataset-name
  • Advanced Settings → OAuth Client ID: leave blank
  • Advanced Settings → OAuth Client Secret: leave blank

Replace your-export-project-id and your-sobjects-dataset-name with your actual values.

Gemini CLI​

Add following to your ~/.gemini/settings.json:

{
  "mcpServers": {
    "cloudaware-mcp": {
      "httpUrl": "https://inbound.prod.cloudaware.com/mcp",
      "headers": {
        "X-CA-ExportProject": "your-export-project-id",
        "X-CA-SObjectsDataset": "your-sobjects-dataset-name"
      }
    }
  }
}

Start Gemini CLI and run /mcp auth cloudaware-mcp to authenticate.

LibreChat​

Add the following to your librechat.yaml:

mcpServers:
  cloudaware-mcp:
    type: streamable-http
    url: https://inbound.prod.cloudaware.com/mcp?exportProject=your-export-project-id&sObjectsDataset=your-sobjects-dataset-name
    timeout: 30000
    serverInstructions: true