Snowflake MCP Server

Execute Query

Execute a SQL query synchronously in Snowflake. Returns results immediately for queries that complete within 45 seconds. Use this for quick queries, data exploration, and interactive analysis.

Execute Query Async

Submit a SQL query for asynchronous execution in Snowflake. Returns a statement handle immediately. Use this for long-running queries, large data operations, or when you need to execute multiple queries in parallel.

Check Query Status

Check the execution status of an asynchronous query. Returns the current state (running, success, failed) and progress information. Use after submitting an async query to monitor its progress.

Get Query Results

Retrieve the results of a completed asynchronous query. Returns the data rows and metadata. Use this after an async query has completed successfully.

Cancel Query

Cancel a running query in Snowflake. Use this to stop long-running queries that are no longer needed or were submitted by mistake.

List Databases

List all databases accessible to the current user in Snowflake. Returns database names, owners, and metadata. Useful for discovering available data sources.

List Schemas

List all schemas in a specific database. Returns schema names, owners, and metadata. Essential for navigating database structure and finding relevant data.

List Tables

List all tables and views in a schema. Returns table names, types (TABLE/VIEW), row counts, and metadata. Critical for data discovery and understanding available datasets.

Describe Table

Get detailed information about a table's structure including column names, data types, nullability, defaults, and keys. Essential for understanding table schema before querying.

Get Table Sample

Retrieve sample rows from a table for quick data exploration. Returns a limited number of rows to understand data format and content without reading the entire table.

Search Objects

Search for tables, views, and other database objects by name pattern across all accessible databases and schemas. Useful for finding specific datasets when you don't know their exact location.

List Warehouses

List all virtual warehouses accessible to the current user. Returns warehouse names, sizes, states (STARTED/SUSPENDED), and configuration. Essential for warehouse management and cost optimization.

Describe Warehouse

Get detailed information about a specific warehouse including size, auto-suspend settings, auto-resume settings, state, and resource configuration. Use this to check warehouse status before executing queries.

Start Warehouse

Resume or start a suspended warehouse. The warehouse will be available for query execution. Use this before running queries if the warehouse is suspended.

Suspend Warehouse

Suspend a running warehouse to stop compute costs. The warehouse will not consume credits while suspended. Use this to optimize costs when the warehouse is not needed.

Get Query History

View recent query execution history with details like query text, execution time, bytes scanned, and status. Useful for debugging, auditing, and understanding query performance.

Get Current Session

Get information about the current session including current database, schema, warehouse, role, and user. Useful for understanding the execution context.

Get Account Usage

Get account-level usage statistics including warehouse credit consumption, storage usage, and data transfer metrics. Essential for cost monitoring and optimization.

Get Row Count

Get the total number of rows in a table. Fast way to check table size without scanning the entire table. Useful for understanding data volume.

Create Table from Query

Create a new table from the results of a SELECT query (CREATE TABLE AS SELECT). Useful for materializing query results, creating derived datasets, or saving analysis results.

Query Information Schema

Query the INFORMATION_SCHEMA to get metadata about databases, schemas, tables, columns, and other objects. Powerful tool for programmatic schema discovery and metadata analysis.

List Stages

List all stages (internal/external storage locations) in a schema. Stages are used for loading and unloading data files. Returns stage names, types, and configuration.

Describe Stage

Get detailed information about a specific stage including type (internal/external), URL, credentials, encryption settings, and file format options.

List Stage Files

List files in a stage location. Shows file names, sizes, modification timestamps, and MD5 checksums. Essential for data loading operations and file management.

Remove Stage Files

Remove files from a stage. Use this to clean up after data loading operations or manage stage storage. Supports pattern matching to remove multiple files.

List Roles

List all roles accessible to the current user. Roles control access to Snowflake objects and operations. Returns role names, owners, and assigned privileges.

Show Grants on Role

Display all privileges granted to a specific role. Shows what databases, schemas, tables, warehouses, and other objects the role can access.

Show Grants on Object

Display all privileges granted on a specific object (database, schema, table, view, etc.). Shows which roles have access to the object and what operations they can perform.

List Users

List all users in the Snowflake account. Returns usernames, login information, default roles, and account status. Requires appropriate privileges.

Describe User

Get detailed information about a specific user including properties like default role, default warehouse, email, login name, display name, and account settings.

List Semantic Views

List all semantic views in a schema. Semantic views provide business-friendly metadata layers over your data, enabling natural language queries through Cortex Analyst.

Describe Semantic View

Get detailed information about a semantic view including its semantic model definition, base tables, metrics, dimensions, and relationships.

Query Semantic View

Query a semantic view using natural language or structured queries. The semantic layer translates business terms to SQL, making data exploration intuitive.

Get Semantic View Metadata

Retrieve metadata about a semantic view's metrics and dimensions. Shows available business metrics, their aggregations, and dimensional attributes for filtering and grouping.

Cortex Search Query

Query a Cortex Search service for semantic/vector search over unstructured data. Used in RAG (Retrieval Augmented Generation) applications to find relevant documents, text chunks, or embeddings based on natural language queries.

Cortex Analyst Query

Ask natural language questions against structured data using Cortex Analyst. Automatically generates SQL from your question using semantic models/views, executes it, and returns results with explanations.

List Cortex Search Services

List all Cortex Search services in a schema. Shows available search services configured for semantic/vector search over your unstructured data.

List Compute Pools

List all compute pools in the account. Compute pools provide containerized compute resources for Snowpark Container Services. Shows pool names, sizes, states, and configuration.

Describe Compute Pool

Get detailed information about a specific compute pool including instance family, size, min/max nodes, auto-resume, auto-suspend settings, and current state.

Suspend Compute Pool

Suspend a compute pool to stop consuming credits. Use this when containerized workloads are not needed. The pool can be resumed later without reconfiguration.

Copy Into Table

Load data from files in a stage into a table. Supports CSV, JSON, Parquet, Avro, ORC, and XML formats. Essential for ETL operations and data ingestion pipelines.

Create View

Create a new view (virtual table) based on a SELECT query. Views provide abstraction layers over tables and can simplify complex queries for end users.

Show File Formats

List all file formats in a schema. File formats define how to parse files during data loading (CSV delimiters, JSON structure, Parquet compression, etc.).