mcp/cockroachdb repository overview

⁠CockroachDB MCP Server

Enable AI agents to manage, monitor, and query CockroachDB using natural language. Perform complex database operations, cluster management, and query execution seamlessly through AI-driven workflows. Integrate effortlessly with MCP clients for scalable and high-performance data operations.

What is an MCP Server?⁠

⁠MCP Info

⁠Image Building Info

Attribute	Details
Dockerfile	https://github.com/amineelkouhen/mcp-cockroachdb/blob/5c323e49ef52dbf1f3d4714dcd2dffa577d368fe/Dockerfile⁠
Commit	5c323e49ef52dbf1f3d4714dcd2dffa577d368fe
Docker Image built by	Docker Inc.
Docker Scout Health Score
Verify Signature	COSIGN_REPOSITORY=mcp/signatures cosign verify mcp/cockroachdb --key https://raw.githubusercontent.com/docker/keyring/refs/heads/main/public/mcp/latest.pub
Licence	MIT License

⁠Available Tools (29)

⁠Tools Details

⁠Tool: analyze_performance

Analyze query performance statistics for a given query or time range.

⁠Tool: analyze_schema

Analyze the schema and provide a summary of tables, views, and relationships.

⁠Tool: bulk_import

Bulk import data into a table from a file (CSV or Avro) stored in cloud or web storage. Supports S3, Azure Blob, Google Storage, HTTP/HTTPS URLs.

⁠Tool: connect

Connect to the default CockroachDB database and create a connection pool.

Returns: A success message or an error message.

⁠Tool: connect_database

Connect to a CockroachDB database and create a connection pool.

⁠Tool: create_database

Enable the creation of new databases.

⁠Tool: create_index

Create a new index on a specified table to improve query performance. This tool allows users to define indexes on one or more columns, enabling faster data retrieval and optimized execution plans for read-heavy workloads.

⁠Tool: create_table

Enable the creation of new tables in the current database. You can instruct the AI to define table names, columns, and their types, streamlining database setup and schema evolution directly through natural language.

⁠Tool: create_view

Create a view from a specific query.

⁠Tool: describe_table

Provide detailed schema information, column definitions, data types, and other metadata for a specified table. This allows the AI to accurately interpret table structures and formulate precise queries or data manipulation commands.

⁠Tool: drop_database

Drop an existing database.

⁠Tool: drop_index

Drop an existing index.

⁠Tool: drop_table

Facilitate the deletion of existing tables from the database. This tool is useful for cleaning up test environments or managing schema changes, always with the necessary confirmations for security.

⁠Tool: drop_view

Drop an existing view.

⁠Tool: execute_query

Execute a SQL query with optional parameters and formatting.

⁠Tool: execute_transaction

Execute a list of SQL queries as a single transaction.

⁠Tool: explain_query

Return CockroachDB's statement plan for a preparable statement. You can use this information to optimize the query. If you run it with Analyze, it executes the SQL query and generates a statement plan with execution statistics.

⁠Tool: get_active_connections

List active connections/sessions to the current database.

Returns: Active sessions on the cluster.

⁠Tool: get_cluster_status

Get cluster health and node distribution.

⁠Tool: get_connection_status

Get the current connection status and details.

Returns: The connection status or an error message.

⁠Tool: get_database_settings

Retrieve current database or cluster settings.

Returns: All cluster settings.

⁠Tool: get_query_history

Get the history of executed queries.

⁠Tool: get_replication_status

Get replication and distribution status for a table or the whole database.

⁠Tool: get_table_relationships

Get foreign key relationships for a table or all tables.

⁠Tool: list_databases

List all databases in the CockroachDB cluster.

Returns: A list of databases with row count or an error message.

⁠Tool: list_tables

List all tables present in the connected Cockroach database instance. This is invaluable for AI to understand the database’s landscape and identify relevant data sources for a given query.

⁠Tool: list_views

List all views in a schema.

⁠Tool: show_running_queries

Show currently running queries on the cluster.

⁠Tool: switch_database

Switch the connection to a different database.

⁠Use this MCP Server

{
  "mcpServers": {
    "cockroachdb": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "CRDB_HOST",
        "-e",
        "CRDB_PORT",
        "-e",
        "CRDB_DATABASE",
        "-e",
        "CRDB_USERNAME",
        "-e",
        "CRDB_SSL_MODE",
        "-e",
        "CRDB_SSL_CA_PATH",
        "-e",
        "CRDB_SSL_KEYFILE",
        "-e",
        "CRDB_SSL_CERTFILE",
        "-e",
        "CRDB_PWD",
        "mcp/cockroachdb"
      ],
      "env": {
        "CRDB_HOST": "127.0.0.1",
        "CRDB_PORT": "26257",
        "CRDB_DATABASE": "defaultdb",
        "CRDB_USERNAME": "root",
        "CRDB_SSL_MODE": "disable",
        "CRDB_SSL_CA_PATH": "",
        "CRDB_SSL_KEYFILE": "",
        "CRDB_SSL_CERTFILE": "",
        "CRDB_PWD": ""
      }
    }
  }
}

Why is it safer to run MCP Servers with Docker?⁠