MCP Usage

Spanner supports Model Context Protocol (MCP) to connect your database to LLM agents.

You can use pre-built tools provided by Spanner to connect your IDE or application to Spanner using MCP.

For detailed instructions, see the official documentation: Connect your IDE to Spanner using MCP

The Spanner MCP server exposes the following tools to LLM agents:

get_instance: Get information about a specific Spanner instance.
list_instances: List all Spanner instances in a given project.
list_configs: List instance configurations available in a project.
get_config: Get information about a specific instance configuration.
create_instance: Create a new Spanner instance.
update_instance: Update a Spanner instance.
create_database: Create a database in a given instance.
get_database_ddl: Retrieve the database schema (DDL).
list_databases: List all databases in a given instance.
create_session: Create a session in a database for query execution.
execute_sql: Execute a SQL statement (both DQL and DML supported).
execute_sql_readonly: Execute a SQL query statement in a single-use read-only transaction.
commit: Commit a transaction.
update_database_schema: Update the schema for a given database.
get_operation: Get the status of a long-running operation.

For more details, see the Spanner MCP reference.

Third-party checks

No alerts10d3 checks · Risk SAFE

Gen Agent Trust Hub10d
This skill is designed to assist with Google Cloud Spanner administration and development. It provides standard workflows for schema design, SQL querying, and resource management using official Google tools and libraries. It includes appropriate safety guardrails requiring user confirmation for destructive database operations.
Socket10d
No alerts
Snyk10d
Risk: LOW · No issues

/spanner-basics