SchemaCrawler AI

SchemaCrawler AI MCP Server

The SchemaCrawler AI MCP Server enables natural language interaction with your database schema using an MCP client in "Agent" mode. It allows users to explore tables, columns, foreign keys, triggers, stored procedures and more simply by asking questions like "Explain the code for the interest calculation stored procedure". You can also ask it to help with SQL, since it knows your schema. This is ideal for developers, DBAs, and data analysts who want to streamline schema comprehension and query development without diving into dense documentation.

What is an MCP Server?

MCP Info

Image Building Info

Available Tools (8)

Tools Details

Tool: database_server_information

Provides database environment and server configuration metadata, including engine type and version, collation, encoding, parameters, capabilities, and platform details. Adapts output to the specific database (such as Oracle, SQL Server, PostgreSQL and so on) to support platform-aware SQL generation and schema analysis.

Tool: describe_entities

Generates detailed documentation for entities in the ER model, including entity type such as strong, weak and subtype entities, and attributes. Supports regex-based entity name filtering to optimize tool performance. Returns data as a JSON object.

entity_name|string optional|Name of entity to describe, from the ER model. May be specified as a regular expression, matching the fully qualified entity name (including the schema). Try not to match all entities, but instead use a regular expression to match a subset or match a single entity, since otherwise results may be large.

This tool is read-only. It does not modify its environment.

Tool: describe_relationships

Generates detailed documentation for relationships in the ER model, including 1..1, 1..M, M..N and optional relationships. Supports regex-based relationship name filtering to optimize tool performance. Returns data as a JSON object.

relationship_name|string optional|Name of relationships to describe, from the ER model. May be specified as a regular expression, matching the fully qualified relationship name (including the schema). Try not to match all relationships, but instead use a regular expression to match a subset or match a single relationships, since otherwise results may be large.

This tool is read-only. It does not modify its environment.

Tool: describe_routines

Generates detailed documentation for database routines (stored procedures and functions), including parameter metadata (input/ output parameters, data types, default values), return types, dependencies (on tables, views, or other routines), and full DDL definitions. Supports regex-based routine name filtering and configurable detail levels to optimize tool performance. Returns data as a JSON object.

routine_name|string optional|Name of database routine (stored procedure or function) to describe. May be specified as a regular expression matching the fully qualified stored procedure or function names (including the schema). Try not to match all routines, but instead use a regular expression to match a subset, since otherwise results may be large.

This tool is read-only. It does not modify its environment.

Tool: describe_tables

Generates detailed documentation for database tables and views, including column definitions (names, data types, constraints, nullability), primary and foreign key relationships, index and trigger information, table attributes, and complete DDL definitions. Supports regex-based table name filtering and configurable detail levels to optimize tool performance. Returns data as a JSON object.

table_name|string optional|Name of database table or view to describe. May be specified as a regular expression, matching the fully qualified table name (including the schema). Try not to match all tables, but instead use a regular expression to match a subset or match a single table, since otherwise results may be large.

This tool is read-only. It does not modify its environment.

Tool: list

Makes an inventory of database schema objects such as tables, views, stored procedures, functions, sequences, synonyms, and more. This is an essential starting point for database exploration, database asset management, and schema analysis. The tool supports object type filtering and pattern-based searching. Returns JSON data.

database_object_type|string optional|Type of database object to list, like tables (including views), routines (that is, stored procedures and functions), schemas (that is, databases or catalogs), sequences, or synonyms. If the parameter is not provided, all database objects are listed.

This tool is read-only. It does not modify its environment.

Tool: list_across_tables

Makes an inventory of table-dependent objects across the database schema, including columns, indexes, foreign keys, triggers, and constraints. Enables impact analysis, dependency tracking, performance tuning, and refactoring without inspecting individual tables. The corresponding tables are identified, and their details can be obtained later by describing those tables. Supports regex-based table name filtering and configurable detail levels to optimize tool performance. Returns JSON data.

dependant_object_name|string optional|Name of table dependant object. May be a regular expression, matching the fully qualified dependant object name (including the schema and table). May match more than one dependant object. Use an empty string if all dependant objects are requested. If not specified, all table dependant objects will be returned, but the results could be large.

table_name|string optional|Name of database table for which dependant objects are described. May be a regular expression, matching the fully qualified table name (including the schema), in which case, multiple tables may be returned. Use an empty string if all tables are requested. If not specified, all tables will be returned, but the results could be large.

This tool is read-only. It does not modify its environment.

Tool: mcp-server-health

Gets the SchemaCrawler AI MCP Server version and uptime status.

This tool is read-only. It does not modify its environment.

Use this MCP Server

{
  "mcpServers": {
    "schemacrawler-ai": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "SCHCRWLR_JDBC_URL",
        "-e",
        "SCHCRWLR_SERVER",
        "-e",
        "SCHCRWLR_HOST",
        "-e",
        "SCHCRWLR_PORT",
        "-e",
        "SCHCRWLR_DATABASE",
        "-e",
        "SCHCRWLR_INFO_LEVEL",
        "-e",
        "SCHCRWLR_LOG_LEVEL",
        "-e",
        "SCHCRWLR_DATABASE_USER",
        "-e",
        "SCHCRWLR_DATABASE_PASSWORD",
        "-v",
        "/local-directory:/local-directory",
        "mcp/schemacrawler-ai"
      ],
      "env": {
        "SCHCRWLR_JDBC_URL": "12.12.12.12",
        "SCHCRWLR_SERVER": "12.12.12.12",
        "SCHCRWLR_HOST": "12.12.12.12",
        "SCHCRWLR_PORT": "9999",
        "SCHCRWLR_DATABASE": "public",
        "SCHCRWLR_INFO_LEVEL": "standard",
        "SCHCRWLR_LOG_LEVEL": "ALL",
        "SCHCRWLR_DATABASE_USER": "SA",
        "SCHCRWLR_DATABASE_PASSWORD": "super_S3cret"
      }
    }
  }
}

Why is it safer to run MCP Servers with Docker?