What features does GibsonAI MCP server provide?

Q: What features does GibsonAI MCP server provide?

get_projects: List all existing GibsonAI projects associated with the authenticated user. Useful when a user refers to a project by name but the UUID is needed; if a .gibsonai file exists, it should be used.. create_project: Create a new GibsonAI project. Checks for an existing .gibsonai file or similar project names before creation and prompts the user to update or create the file with the new UUID.. get_project_details: Fetch metadata and configuration for a given project using its UUID, ideal for loading project-specific context from an existing .gibsonai file.. get_project_hosted_database_details: Return credentials, connection string, dialect, and other details needed to query the hosted GibsonAI database.. update_project: Rename a project by updating its UUID-linked name. Currently supports only the project_name field.. submit_data_modeling_request: Submit natural-language data modeling requests. The internal modeler handles the request end-to-end instead of manual schema design.. deploy_project: Trigger automatic schema migrations and deploy the current schema to all GibsonAI-supported databases.. query_database: Run SQL queries against a database using the API key associated with that database, with proper quoting per SQL dialect.

GibsonAI MCP Server 2025: Features & Setup Guide

Overview

The GibsonAI MCP Server provides a set of tools to MCP Clients like Cursor, Windsurf, or Claude Desktop. These clients can use these tools to interact with GibsonAI projects and databases using natural language instructions. You can accomplish tasks with GibsonAI directly in your favorite IDE or chat interface, including creating new GibsonAI projects and designing database schemas, viewing project structure, schema diagrams, and a summary of tables and relationships. It supports applying schema changes and triggering automatic migrations, running SQL queries against your database, and deploying projects to development or production environments. It can also seed tables with mock data and help build full-stack apps. To use the server, you must log in to the Gibson CLI before the MCP server will work. The documentation provides setup instructions for Cursor, Windsurf, Claude Desktop, VS Code, and Cline, demonstrating how to configure MCP servers across popular tools. A note on distribution clarifies that the MCP server code lives in the Gibson CLI, enabling shared authentication and API interaction logic for faster feature delivery. This MCP server aims to streamline AI-assisted database design, querying, migration, and deployment from within editors or chat interfaces.

Features

get_projects

List all existing GibsonAI projects associated with the authenticated user. Useful when a user refers to a project by name but the UUID is needed; if a .gibsonai file exists, it should be used.

create_project

Create a new GibsonAI project. Checks for an existing .gibsonai file or similar project names before creation and prompts the user to update or create the file with the new UUID.

get_project_details

Fetch metadata and configuration for a given project using its UUID, ideal for loading project-specific context from an existing .gibsonai file.

get_project_hosted_database_details

Return credentials, connection string, dialect, and other details needed to query the hosted GibsonAI database.

update_project

Rename a project by updating its UUID-linked name. Currently supports only the project_name field.

submit_data_modeling_request

Submit natural-language data modeling requests. The internal modeler handles the request end-to-end instead of manual schema design.

deploy_project

Trigger automatic schema migrations and deploy the current schema to all GibsonAI-supported databases.