databricks

What This Skill Does

The Databricks skill for OpenClaw provides a powerful bridge to your Databricks workspace, enabling seamless control over your data engineering and analytics infrastructure directly from your chat interface. By leveraging the Membrane CLI and integration framework, this skill automates the complexity of authentication and API management. Whether you need to manage compute resources like SQL Endpoints and Spark Clusters, trigger automated data pipelines via Jobs, or execute specific Notebooks, this skill handles the heavy lifting. It acts as an orchestrator for your data environment, allowing you to monitor and adjust your infrastructure without switching contexts to the Databricks UI.

Installation

To begin, ensure you have the OpenClaw environment ready. First, install the skill using the command: clawhub install openclaw/skills/skills/gora050/databricks. Once the skill is installed, you must install the Membrane CLI via npm install -g @membranehq/cli. Perform the initial setup by running membrane login --tenant. If you are working in a headless environment, follow the manual URL authentication flow. Next, identify the correct Databricks connector using membrane search databricks --elementType=connector --json and initiate your session with membrane connect --connectorId=YOUR_ID. You can verify active sessions at any time using membrane connection list --json.

Use Cases

This skill is ideal for data engineers who need to quickly spin up or shut down compute clusters to optimize costs, data analysts requiring immediate SQL endpoint availability for dashboarding, and DevOps teams managing automated ETL workflows. It is particularly effective for troubleshooting, as users can list jobs and check statuses in real-time, or trigger ad-hoc notebook executions to validate data quality following a failed pipeline run.

Example Prompts

1. "OpenClaw, list all my active Databricks clusters and tell me which ones are currently running."
2. "Please start the SQL Endpoint named 'production-analytics' so I can run my end-of-month reports."
3. "Trigger the daily ingestion job in my Databricks workspace and let me know when it finishes."

Tips & Limitations

Always check for existing connections before creating a new one to avoid cluttering your profile. Use membrane action list frequently to discover newly updated capabilities as your Databricks environment evolves. Note that this skill requires appropriate permissions within your Databricks Workspace to execute actions like starting or stopping clusters. Ensure your API tokens or OAuth scopes are correctly configured to prevent unauthorized access denials during runtime. Always monitor your cloud provider's billing dashboard if you are frequently starting and stopping large compute resources.