Skip to main content

MCP server for Databricks integration

Project description

Databricks Unity Catalog View Access

Script om Databricks Unity Catalog via MCP te benaderen.

Setup Stappen

0. Zorg dat je een Python environment hebt

Python 3.12.11 werkt.

1. Installeer Dependencies

pip install -r requirements.txt

2. Databricks Configuratie (Vereist voor externe toegang)

A. Genereer Personal Access Token

  1. Log in op Databricks workspace: https://adb-4268269233410651.11.azuredatabricks.net
  2. Klik rechtsboven op je gebruikersnaam → Settings
  3. Ga naar DeveloperAccess tokens
  4. Klik Generate new token
  5. Geef een naam (bijv. "Python Script Access") en optioneel een expiration
  6. Kopieer de token (wordt maar 1x getoond!)

B. Verkrijg SQL Warehouse HTTP Path

  1. Ga in Databricks naar SQL Warehouses (via zijmenu)
  2. Selecteer een actieve SQL Warehouse (of maak een nieuwe aan)
  3. Klik op de warehouse en ga naar Connection Details
  4. Kopieer de HTTP Path (format: /sql/1.0/warehouses/xxxxx)

C. Verkrijg Cluster ID (Voor Python Code Execution)

Als je de execute_python_code tool wilt gebruiken:

  1. Ga in Databricks naar Compute (via zijmenu)
  2. Selecteer een actieve cluster (of maak een nieuwe aan)
  3. Kopieer de Cluster ID uit de URL of cluster details

Let op: De Command Execution API werkt alleen met classic all-purpose clusters, niet met serverless compute.

D. Configureer Environment Variables

Maak een .env file aan conform:

DATABRICKS_ACCESS_TOKEN=<je access token>
DATABRICKS_HTTP_PATH=/sql/1.0/warehouses/15c864e4e063e48c
DATABRICKS_SERVER_HOSTNAME=adb-4268269233410651.11.azuredatabricks.net
DATABRICKS_CLUSTER_ID=<je cluster id voor python execution>

Let op: DATABRICKS_CLUSTER_ID is alleen vereist als je de execute_python_code tool wilt gebruiken.

5. Permissies Check

Zorg dat je Databricks account heeft:

  • USE CATALOG rechten op main catalog
  • USE SCHEMA rechten op ralf schema
  • SELECT rechten op full_view_with_location view

Test met:

SHOW GRANTS ON VIEW main.ralf.full_view_with_location;

Publishing to PyPI

This project is automatically published to PyPI using GitHub Actions when you push a version tag.

How to Publish a New Version

  1. Update the version number in pyproject.toml
  2. Commit your changes and create a tag:
    # Update version in pyproject.toml naar bijv. 0.1.2
    git add pyproject.toml
    git commit -m "Bump version to 0.1.2"
    git tag v0.1.2
    git push origin v0.1.2
    
  3. The GitHub Action will automatically build and publish to PyPI

Setup Requirements

Before the first publish, you need to add your PyPI API token to GitHub Secrets:

  1. Go to your repository on GitHub
  2. Navigate to SettingsSecrets and variablesActions
  3. Click New repository secret
  4. Name: PYPI_API_TOKEN
  5. Value: Your PyPI API token (starts with pypi-)
  6. Click Add secret

The workflow will automatically trigger when you push a tag matching the pattern v*.*.* (e.g., v0.1.2, v1.0.0).

Troubleshooting

Fout: "Invalid access token"

  • Check of DATABRICKS_TOKEN environment variable correct is gezet
  • Genereer een nieuwe token als deze verlopen is

Fout: "Warehouse not found"

  • Controleer of HTTP Path correct is ingevuld
  • Zorg dat SQL Warehouse actief is (Status: Running)

Fout: "Table or view not found"

  • Verifieer permissies via Databricks UI
  • Check of catalog/schema/view namen exact kloppen (case-sensitive!)

Fout: "Connection timeout"

  • Check internet connectie
  • Verifieer IP whitelisting in Databricks
  • Test VPN verbinding indien van toepassing

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

databricks_mcp_tools-0.1.2.tar.gz (26.7 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

databricks_mcp_tools-0.1.2-py3-none-any.whl (41.2 kB view details)

Uploaded Python 3

File details

Details for the file databricks_mcp_tools-0.1.2.tar.gz.

File metadata

  • Download URL: databricks_mcp_tools-0.1.2.tar.gz
  • Upload date:
  • Size: 26.7 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.1.0 CPython/3.13.7

File hashes

Hashes for databricks_mcp_tools-0.1.2.tar.gz
Algorithm Hash digest
SHA256 4e077253e55f1cf1e701bcd1ef21e5f8d1744bccc364334c2f713c779fb7e0da
MD5 c0ba1785f9e69cc142799485afcf0ada
BLAKE2b-256 a156ba8e5eae9811d4e2839e0242d19ab4cd350213a2286d7a292c629b4a177e

See more details on using hashes here.

File details

Details for the file databricks_mcp_tools-0.1.2-py3-none-any.whl.

File metadata

File hashes

Hashes for databricks_mcp_tools-0.1.2-py3-none-any.whl
Algorithm Hash digest
SHA256 86e8a4c3b95b201f136dd39026b9b985010ccaf76c933bd88dabff9db8cb07bf
MD5 134aadca5177f5040e55ad8532d95ec0
BLAKE2b-256 841cbcb420cf698c8d62104f63fd36c4987bbc6ae8dd97bd2c99785a17d1633f

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page