Overview

Run notebooks and read lakehouse tables under governance.

  • · Vendor-supported OAuth or scoped service credentials
  • · Read and write access configured separately
  • · Tool calls and outcomes recorded per run
Connector contract

Control what agents can do in Databricks.

Exact authentication methods and available scopes depend on your Databricks plan and deployment. Context keeps the connector configuration explicit.

Read

Query the governed datasets and views selected for the runbook.

Act

Materialize approved results to a designated table or workspace.

Verify

Keep query inputs and returned records attached to the run trace.

Set up the connection

  1. 01

    Authenticate

    Connect Databricks with an admin-approved OAuth flow or scoped credential supported by the vendor.

  2. 02

    Choose scope

    Select the accounts, workspaces, projects, or resources available to the runbook.

  3. 03

    Map policy

    Set read and write permissions, reviewer gates, and the identity used for each action.

  4. 04

    Test a run

    Verify one read, one permitted action, and the resulting audit record before rollout.

Connect Databricks to Context.

Request demo