Singlestore

Singlestore MCP Server

MCP server for interacting with SingleStore Management API and services.

What is an MCP Server?

MCP Info

Image Building Info

Available Tools (28)

Tools Details

Tool: create_cloud_function

Deploy a notebook as a cloud function.

The notebook must already exist in the Stage file system shared space. Use
stage_upload_file to upload a notebook first, then deploy it as a cloud function.

The notebook must define a FastAPI app, and initialize it using a singlestore tool.
Base example:
    from fastapi import FastAPI
    import singlestoredb.apps as apps
    app = FastAPI()
    @app.get("/")
    async def root():
        return {"message": "hello world"}
    await apps.run_function_app(app)

Target_type is optional, but if it specifies a virtual workspace, it must also specify a database_name.

Tool: create_job_from_notebook

Create a scheduled job to run a notebook (uploaded to shared space).

Tool: create_notebook_file

Create a Jupyter notebook file in the correct singlestore format and saves it to a temporary location.

This tool validates the provided content against the Jupyter notebook schema and creates a properly
formatted .ipynb file in a temporary location. The content is converted from the simplified format
to the full Jupyter notebook format.

Each code cell MUST include a "language" field in its metadata for SingleStore to
properly recognize the cell. If not provided or empty, it defaults to "python".
Supported languages are "python" and "sql"; other values are rejected.

Tool: create_starter_workspace

Create a new starter workspace using the SingleStore SDK.

This tool provides a modern SDK-based approach to creating starter workspaces,
offering improved reliability and better error handling compared to direct API calls.

Tool: delete_cloud_function

Delete a cloud function by its ID.

Tool: delete_job

Delete a scheduled job by its ID.

Tool: get_cloud_function

Get detailed information about a specific cloud function including its status, endpoint, and configuration.

Tool: get_cloud_function_token

Get an authentication token for invoking a cloud function.

Returns a JWT token and its expiration time. Use this token in the
Authorization header when making HTTP requests to the cloud function's
endpoint.

Tool: get_job

Retrieve details of a scheduled job by its ID.

Tool: get_user_info

Retrieve all information about the current user.

Returns:
    dict: User information including userID, email, firstName, lastName.

Performance Tip:
Cache the returned info when making multiple API calls.

Tool: list_cloud_functions

List all cloud functions in the current organization.

Returns name, status, endpoint, and description for each cloud function.

Tool: list_regions

List all available deployment regions where SingleStore workspaces can be deployed by the user.

Returns region information including:
- regionID: Unique identifier for the region
- provider: Cloud provider (AWS, GCP, or Azure)
- name: Human-readable region name (e.g., Europe West 2 (London), US West 2 (Oregon))

Use this tool to:
1. Select optimal deployment regions based on:
   - Geographic proximity to users
   - Compliance requirements
   - Cost considerations
   - Available cloud providers
2. Plan multi-region deployments

Tool: list_sharedtier_regions

List all regions where shared tier workspaces can be created.

This tool provides information about available regions for creating starter workspaces,
including region names and cloud providers.

Tool: list_starter_workspaces

List all starter (virtual) workspaces available to the user in SingleStore.

Returns detailed information about each starter workspace:
- virtualWorkspaceID: Unique identifier for the workspace
- name: Display name of the workspace
- endpoint: Connection endpoint URL
- databaseName: Name of the primary database
- mysqlDmlPort: Port for MySQL protocol connections
- webSocketPort: Port for WebSocket connections
- state: Current status of the workspace

Use this tool to:
1. Get starter workspace IDs for other operations
2. Check starter workspace availability and status
3. Obtain connection details for database access

Tool: organization_info

Retrieve information about the current user's organization in SingleStore.

Returns organization details including:
- orgID: Unique identifier for the organization
- name: Organization display name

Tool: resume_workspace

Resume a workspace within a specified workspace group in SingleStore.

Tool: run_sql

Use this tool to execute a single SQL statement against a SingleStore database.

Generally you should *NOT* specify the username and password when calling this tool.
You must only use those parameters if a previous attempt to call this method failed and
indicated that the `username` and `password` must be specified.
Credentials for a specific workspace and database combination will be cached.

Returns:
- Query results with column names and typed values
- Row count and metadata
- Execution status
- Workspace type ("shared" for starter workspaces, "dedicated" for regular workspaces)
- Workspace name

Tool: stage_create_folder

Create a folder in Stage.

Tool: stage_delete

Delete a file or folder from Stage.

For folders, ensure the path ends with '/'.
For files, the path must not end with '/'.

Tool: stage_get_file

Get a file from Stage by path. Returns metadata, a download URL, or text content.

Recommended workflow: first call with return_type='metadata' to check file size
and type, then decide whether to fetch the full content or just the URL.

Tool: stage_list_files

List files and folders in a Stage deployment's file system.

Lists the contents of the root folder or a specific subfolder in the Stage
file system attached to a SingleStore deployment (workspace group or starter
workspace).

Tool: stage_move

Move or rename a file or folder in Stage. For folders, ensure the path ends with '/'. For files, the path must not end with '/'.

Works like the `mv` command - can rename and/or move into a different folder.

Tool: stage_upload_file

Upload a file to Stage from text content or a local file path.

Provide exactly one of `content` or `local_path`:
- content: Text content to upload as the file body.
- local_path: Absolute path to a local file to upload.

Tool: terminate_starter_workspace

Permanently delete a starter workspace in SingleStore with safety confirmations.

⚠️  WARNING: This action CANNOT be undone. All workspace data will be permanently lost.
Make sure to backup important data before proceeding.

Safety Features:
- Requires explicit user confirmation (if elicitation is supported)
- Validates workspace existence
- Provides warning messages
- Includes error handling

Tool: update_cloud_function

Update a cloud function's configuration. Only provided fields are updated.

Tool: upload_notebook_file

Upload a notebook file from a local local_path to SingleStore shared or personal space.

This tool validates the notebook schema before uploading. If upload_name or upload_location
are not provided, the user will be prompted through elicitation.

Tool: workspace_groups_info

List all workspace groups accessible to the user in SingleStore.

Returns detailed information for each group:
- workspaceGroupID: Unique identifier for the group
- name: Display name of the workspace group
- region: Region information (name, provider)
- firewallRanges: List of allowed IP ranges for the group
- allowAllTraffic: Whether all traffic is allowed to the group
- createdAt: Timestamp of group creation
- terminatedAt: Timestamp when the group was terminated (if applicable)


Use this tool to:
1. Get workspace group IDs for other operations
2. Plan maintenance windows

Related operations:
- Use workspaces_info to list workspaces within a group
- Use execute_sql to run queries on workspaces in a group

Tool: workspaces_info

List all workspaces within a specified workspace group in SingleStore.

Returns detailed information for each workspace:
- createdAt: Timestamp of workspace creation
- deploymentType: Type of deployment (e.g., 'PRODUCTION')
- endpoint: Connection URL for database access
- name: Display name of the workspace
- size: Compute and storage configuration
- state: Current status (e.g., 'ACTIVE', 'PAUSED')
- terminatedAt: End timestamp if applicable
- workspaceGroupID: Workspacegroup identifier
- workspaceID: Unique workspace identifier

Use this MCP Server

{
  "mcpServers": {
    "singlestore": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "MCP_API_KEY",
        "mcp/singlestore"
      ],
      "env": {
        "MCP_API_KEY": "<MCP_API_KEY>"
      }
    }
  }
}

Why is it safer to run MCP Servers with Docker?