Query the shared GTM Snowflake warehouse through one read-only service account.
Asks Snowflake Cortex Analyst a natural-language question against a semantic model or semantic view. Returns content blocks (text, sql, suggestions). Exactly one of semantic_model_file or semantic_view must be provided. Review the returned SQL, then run it with execute_query.
Queries a Snowflake Cortex Search service for low-latency semantic search over unstructured text. Returns matching results with the requested columns.
Describes the columns of a table via DESCRIBE TABLE. Returns name, type, kind, null?, default, primary key, and unique key for each column.
Executes a read-only SQL query (SELECT, SHOW, DESCRIBE). Write operations (INSERT, UPDATE, DELETE, DDL) are always rejected — this integration is read-only by design. Returns columns, rows (values as strings), row_count, and a truncation warning if results exceed 10,000 rows.
Lists all databases visible to the connection via SHOW DATABASES. Returns name, origin, owner, comment, and creation time.
Lists all schemas in a database via SHOW SCHEMAS. Returns name, database_name, owner, comment, and creation time.
Lists all tables in a schema via SHOW TABLES. Returns name, database_name, schema_name, kind, rows (row count), bytes, and comment.
Lists all warehouses visible to the connection via SHOW WAREHOUSES. Returns name, state (STARTED/SUSPENDED), type, size, and other warehouse metadata.
One-click install from the Simtheory store.
Authorize access to your Snowflake (Team) account.
Your AI can now interact with Snowflake (Team).
Query and explore Amazon Redshift data warehouses - discover clusters, databases, schemas, tables, and execute SQL queries.
Execute Python code in a sandboxed environment. Analyze data, create charts, edit files, and more.
Query your Snowflake data warehouse — execute SQL, explore warehouses, databases, schemas and tables, and use Cortex AI for natural-language analytics and search.
Sign up for Simtheory and install this MCP directly into your workspace.
Get Started