Bauplan MCP Server

A Model Context Protocol (MCP) server that gives AI assistants (Claude Code, Claude Desktop, Cursor) access to Bauplan lakehouse operations: querying tables, schema inspection, branch management, and running pipelines. A video walkthrough demonstrates setup and usage.

Note

This project is released in Beta under MIT license. APIs and features may change without notice as we continue development.

Important

A hosted MCP server is now available for existing Bauplan users at https://mcp.use1.aprod.bauplanlabs.com/mcp - you no longer need to run the server locally. Add it as a custom connector in Claude Desktop (or any MCP client) and authenticate with your Bauplan API key. See Execution modes for all the ways to run the server.

Overview

This repository contains the Bauplan MCP Server — a Model Context Protocol server that gives AI assistants access to Bauplan lakehouse operations. A blog post with context and background is available here.

Tip

Looking for the best local AI setup with Bauplan? Check out BauplanLabs/bauplan-skills — it includes agent playbooks (CLAUDE.md), skills, and everything you need to get AI coding assistants working with Bauplan via CLI and SDK, without running an MCP server.

Execution modes

The server supports three execution modes. They differ in who runs the server and in which tools are exposed: tool visibility is set by FastMCP tags and can be tuned with MCP_VISIBLE_TOOL_TAGS. All three authenticate with your Bauplan API key.

Local - run the server on your own machine and point your AI assistant at http://localhost:8000/mcp; exposes every tool by default. See MCP Quick Start.
Self-hosted - deploy the server on your own infrastructure and point your assistant at your endpoint, passing your Bauplan API key as a bearer token; control the exposed tools with MCP_VISIBLE_TOOL_TAGS. See Bauplan Credentials and Container Runtime.
Bauplan public endpoint - skip running a server and connect directly to Bauplan's hosted endpoint at https://mcp.use1.aprod.bauplanlabs.com/mcp; exposes only the remote-safe tools.

MCP Quick Start

You can get started in one minute with your existing AI assistant: a video setup with Claude Desktop and Claude Code is also available here for reference.

You need:

a Bauplan API key properly configured in your local config file (default profile) - the server will pick it up automatically (see below for alternative authentication methods);
uv (or a standard pip managed virtual environment, see below);
an AI platform able to leverage the MCP, as for example Claude Code, Cursor, Claude Desktop.

Warning

do not use an Admin Bauplan API key: while the server will refuse to write on main, it is good practice to use a non-admin key for AI-assisted development (see our roadmap below for more details on upcoming security features).

Start the server with:

uv sync
uv run python main.py --transport streamable-http

The MCP server is now available at http://localhost:8000/mcp. You can configure the server in Claude Code for example with:

claude mcp add -t http mcp-bauplan http://localhost:8000/mcp

Similar commands can be run on Claude Desktop or Cursor to enable the AI to access the server.

Et voilà! You can now start asking your AI questions about your data lakehouse (and much more!).

Advanced Configurations

Bauplan Credentials

The Beta release covers the local development use case. Authentication to your Bauplan lakehouse happens as follows:

if you do not specify a Bauplan profile as a flag, the default one on the machine running the server will be used at every interaction with the lakehouse.
if you specify a profile as a flag, this profile will be used instead when instantiating a Bauplan client.
if you specify a header in your assistant - either Authorization: Bearer <your-bauplan-api-key> or Bauplan: <your-bauplan-api-key> (e.g. in Claude Code claude mcp add -H "Authorization: Bearer <your-bauplan-api-key>" ...) -, that value will be used instead when instantiating a Bauplan client. This is convenient for quick tests, and opens up the possibility of hosting the catalog on a shared infrastructure, delegating to clients the Bauplan API key management.

For example, if you are connecting to a remotely hosted MCP server that delegates Bauplan authentication to the client, you can register it in Claude Code and pass your own bearer token with:

claude mcp add -t http -H "Authorization: Bearer <your-bauplan-api-key>" mcp-bauplan https://<your-mcp-host>/mcp

Server CLI Options

The server supports the following CLI options, mostly useful for specifying alternative transport options:

Option	Default	Description	Used With
`--transport`	`stdio`	Transport protocol: `stdio`, `sse`, or `streamable-http`	All commands
`--host`	`127.0.0.1`	Host to bind to (localhost by default)	`sse`, `streamable-http` only
`--port`	`8000`	Port to bind to	`sse`, `streamable-http` only
`--profile`	`None`	Bauplan profile to use	All commands

Note: The --host and --port options are ignored when using stdio transport since it communicates through stdin/stdout.

Container Runtime

The Docker image starts HTTP transports with Gunicorn and Uvicorn workers. This is meant for remote deployments where several requests can be handled concurrently. Local MCP usage should continue to use python main.py, especially for stdio.

Useful environment variables:

Variable	Default	Description
`MCP_TRANSPORT`	`streamable-http`	HTTP transport for the container, or `stdio`
`MCP_WORKERS`	`4`	Gunicorn worker count for HTTP transports
`MCP_KEEP_ALIVE`	`65`	Gunicorn keep-alive timeout in seconds
`PORT`	`8000`	Container listen port for HTTP transports
`MCP_AUTH_MODE`	`none`	Use `api-key-oauth` for remote OAuth clients
`MCP_PUBLIC_BASE_URL`	required for OAuth	Public MCP server base URL. Requires `MCP_AUTH_MODE=api-key-oauth`
`MCP_LOG_TOOL_ARGS`	`false`	Log full tool arguments for debugging. Emits a warning when `MCP_PUBLIC_BASE_URL` is set
`MCP_VISIBLE_TOOL_TAGS`	unset locally, `remote` in OAuth mode	Comma-separated allowlist of FastMCP tool tags to expose
`MCP_OAUTH_SECRET`	required for OAuth	Stable secret used to sign tokens and encrypt API keys
`MCP_OAUTH_TRUSTED_REDIRECTS`	Claude and ChatGPT callbacks	Comma-separated trusted redirect list. Supports a trailing `*` for trusted path prefixes

Tool visibility is controlled with FastMCP tags. Local runs expose every tool by default. Remote OAuth runs expose only tools tagged remote unless MCP_VISIBLE_TOOL_TAGS is set explicitly. This keeps local-only tools such as project_run and run_query_to_csv out of shared server deployments. When multiple tags are configured, a tool is visible if it has at least one matching tag.

Useful examples:

# Expose only remote-safe tools.
MCP_VISIBLE_TOOL_TAGS=remote uv run python main.py --transport streamable-http

# Expose only read-only tools for inspection.
MCP_VISIBLE_TOOL_TAGS=read uv run python main.py --transport streamable-http

Current tags are remote, local, read, write, and destructive.

OAuth clients with valid HTTPS redirect URIs can still register dynamically. Redirects outside MCP_OAUTH_TRUSTED_REDIRECTS are shown to users as unverified before they continue.

Claude Desktop

To add the Bauplan MCP server to Claude Desktop, follow the guide to get to your claude_desktop_config.json file.

You can then add this configuration (modify the paths as needed):

{
  "mcpServers": {
    "mcp-bauplan": {
      "command": "/path/to/bauplan-mcp-server/.venv/bin/python3",
      "args": [
        "/path/to/bauplan-mcp-server/main.py",
        "--transport",
        "stdio"
      ],
      "workingDirectory": "/path/to/bauplan-mcp-server/"
    }
  }
}

Quit and restart Claude Desktop. Now all Bauplan tools are available to your assistant, as this video demonstrates.

MCP Inspector

Start the MCP Inspector if you wish to manually test the server (Node is required):

npx @modelcontextprotocol/inspector

Now, configure the inspector with the proper variables, e.g. for Streamable HTTP:

Transport Type: Streamable HTTP
URL: http://localhost:8000/mcp
Session Token: Use the token from inspector output

Features

Tool List

Data Operations

get_tables: List all tables in a branch/namespace
get_table: Get schema for a specific table (more efficient for single table)
run_query: Execute SELECT queries on tables
run_query_to_csv: Execute SELECT queries and save results directly to CSV file (local/non-OAuth servers only, scalar data types only)

Branch Management

get_branches: List branches with optional filters
get_branch: Get a branch by name
get_commits: Get commit history from branches
create_branch: Create new branches from references
merge_branch: Merge branches with custom commit messages
delete_branch: Delete branches (with safety checks)

Namespace Management

get_namespaces: List available namespaces in a branch
get_namespace: Get a namespace by name
create_namespace: Create new namespaces in branches
delete_namespace: Delete namespaces from branches

Tag Management

get_tags: Get tags with optional filters
get_tag: Get a tag by name
create_tag: Create a new tag from a reference
delete_tag: Delete a tag

Table Management

create_table: Create a table from S3 location using schema detection (creates ICEBERG table structure but doesn't populate data)
plan_table_creation: Create a table import plan from S3 location (generates YAML schema plan with job tracking)
apply_table_creation_plan: Apply a table creation plan to resolve schema conflicts (returns job_id for tracking)
delete_table: Delete a table from a specific branch
import_data: Import data into an existing table from S3 location (returns job_id for tracking)
revert_table: Revert a table from a source reference to a target branch with optional replacement

Project Management

project_run: Run a Bauplan project from a specified directory and reference with configurable parameters (local/non-OAuth servers only)
code_run: Run a Bauplan project from code files provided as a dictionary (useful for clients that cannot submit paths), automatically creates temporary directory and validates project structure

Job Management

get_jobs: List jobs in the Bauplan system with optional filtering for all users
get_job: Get detailed information about a specific job by its ID
cancel_job: Cancel a running job by its ID and get updated job status

User Management

get_user_info: Get information about the current authenticated user (username and full name)

Instructions and Guidance

get_instructions: Get detailed instructions for specific Bauplan use cases (pipeline, data, repair, wap, test, sdk)

License

This project is provided with no guarantees under the attached MIT License.

Name		Name	Last commit message	Last commit date
Latest commit History 85 Commits
.github/workflows		.github/workflows
mcp_bauplan		mcp_bauplan
tests		tests
.dockerignore		.dockerignore
.gitignore		.gitignore
.pre-commit-config.yaml		.pre-commit-config.yaml
.python-version		.python-version
Dockerfile		Dockerfile
LICENSE		LICENSE
README.md		README.md
entrypoint.sh		entrypoint.sh
main.py		main.py
mise.toml		mise.toml
pyproject.toml		pyproject.toml
uv.lock		uv.lock

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Repository files navigation

Bauplan MCP Server

Overview

Execution modes

MCP Quick Start

Advanced Configurations

Bauplan Credentials

Server CLI Options

Container Runtime

Claude Desktop

MCP Inspector

Features

Tool List

Data Operations

Branch Management

Namespace Management

Tag Management

Table Management

Project Management

Job Management

User Management

Instructions and Guidance

License

About

Uh oh!

Releases

Packages

Uh oh!

Uh oh!

Contributors

Uh oh!

Languages

Uh oh!

Folders and files

Latest commit

History

Repository files navigation

Bauplan MCP Server

Overview

Execution modes

MCP Quick Start

Advanced Configurations

Bauplan Credentials

Server CLI Options

Container Runtime

Claude Desktop

MCP Inspector

Features

Tool List

Data Operations

Branch Management

Namespace Management

Tag Management

Table Management

Project Management

Job Management

User Management

Instructions and Guidance

License

About

Topics

Resources

License

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Uh oh!

Contributors

Uh oh!

Languages

Packages