MCP Datastore

Provides access to Google Cloud Datastore in Datastore mode with entities, queries, keys, and transactions.

javascript

GitHub Stars

javascript

Language

4 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "petekmet-mcp-gcp-datastore": {
      "command": "node",
      "args": [
        "d:\\projects\\mcp-datastore\\build\\index.js"
      ],
      "env": {
        "GOOGLE_CLOUD_PROJECT": "your-project-id",
        "GOOGLE_APPLICATION_CREDENTIALS": "C:\\\\path\\\\to\\\\service-account-key.json"
      }
    }
  }
}

You can access Google Cloud Firestore in Datastore mode through an MCP server that exposes entity operations, queries, key management, and transactions. This server lets you perform common Datastore actions from a local or remote MCP client, enabling streamlined data management and analytics workflows.

How to use

You interact with the server from an MCP client to perform create, read, update, delete, and query operations on your Datastore entities. Use the available tools to manage entities, run queries, allocate keys, and execute transactional workloops. You can run the server locally and point your MCP client at the local process, or connect to a remote MCP endpoint if you deploy the server elsewhere.

Practical usage patterns include: inserting new entities with structured data, retrieving entities by kind and key, updating existing entities, running filtered queries with ordering, and performing aggregations like counts or sums. You can also allocate IDs for incomplete keys and execute multiple operations inside a single transactional boundary for consistency.

How to install

Prerequisites are minimal: you need Node.js and npm installed on your system. You also require access to a Google Cloud project with the Datastore API and a service account key file with Datastore permissions.

Step 1: Install dependencies and build the server.

npm install
npm run build

Step 2: Configure environment variables. You can set them in your shell or provide them when starting the server.

Step 3: Start the server.

npm start

Configuration and runtime notes

Set the path to your service account JSON key and your Google Cloud project ID so the server can authenticate with Google Cloud Datastore.

export GOOGLE_APPLICATION_CREDENTIALS=/path/to/service-account-key.json
export GOOGLE_CLOUD_PROJECT=your-project-id

If you prefer, you can pass these variables inline when starting the server.

## Claude Desktop integration example

If you want to connect Claude Desktop to run the MCP server locally, you can specify the server as a stdio process with the appropriate environment variables.

{ "mcpServers": { "datastore": { "command": "node", "args": [ "build/index.js" ], "env": { "GOOGLE_APPLICATION_CREDENTIALS": "/path/to/service-account-key.json", "GOOGLE_CLOUD_PROJECT": "your-project-id" } } } }

Note: adjust paths and project IDs to match your environment.

Available tools

datastore_get

Retrieve a single entity by kind and key.

datastore_insert

Insert a new entity into Datastore.

datastore_update

Update fields of an existing entity.

datastore_upsert

Insert or update an entity depending on its existence.

datastore_delete

Delete an entity by kind and key.

datastore_query

Query entities with filters and ordering.

datastore_runAggregationQuery

Execute aggregation queries such as count, sum, or average.

datastore_listKinds

List all entity kinds present in Datastore.

datastore_allocateIds

Allocate IDs for incomplete keys to prepare for insertion.

datastore_createKey

Create a complete or incomplete key for an entity.

datastore_runInTransaction

Run multiple operations within a single transaction.

Built by

VeilStrat

AI signals for GTM teams