Databricks SQL Read-Only for Openclaw

A secure, read-only interface for executing Databricks SQL queries using Openclaw Skills.

kansodata
v0.1.1
Apr 9, 2026
1
943
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install kansodata-databricks

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install kansodata-databricks using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is Databricks SQL Read-Only?

The Databricks integration within the Openclaw Skills ecosystem enables developers to execute read-only SQL queries against their data lakehouse with built-in safety protocols. By leveraging Openclaw Skills, teams can automate data exploration tasks while maintaining strict security boundaries, preventing unauthorized data mutations or multi-statement execution. This skill is ideal for teams that require high-assurance data access for AI agents while ensuring the integrity of their production Databricks environments.

Databricks SQL Read-Only Use Cases

  • Querying large datasets in Databricks for quick analysis via an AI agent.
  • Generating and validating read-only SQL drafts for complex data transformations.
  • Safely exploring data schemas and table structures in dev, staging, or prod environments.
  • Integrating Databricks data retrieval into broader Openclaw Skills automated workflows.

How Databricks SQL Read-Only Works

  1. The user provides essential connection details including the Databricks workspace URL, environment, and warehouse ID.
  2. The skill identifies the target catalog and schema scope to define the query boundary.
  3. The AI agent generates a safe, single-statement SQL draft (using SELECT or WITH... SELECT).
  4. The databricks_sql_readonly tool validates the query against a blacklist of mutating keywords and multi-statement shapes.
  5. The tool executes the validated query and returns the results along with a validation checklist and risk notes.

Databricks SQL Read-Only Setup

To use this skill, ensure your environment variables are configured for your Databricks workspace:

# Configure Databricks access for Openclaw Skills
export DATABRICKS_HOST="https://your-workspace-url.databricks.com"
export DATABRICKS_TOKEN="your-personal-access-token"
export DATABRICKS_WAREHOUSE_ID="your-sql-warehouse-id"

Databricks SQL Read-Only Data Schema & Taxonomy

The skill utilizes the following data structure to manage Databricks interactions:

Input Parameter Description
Workspace URL The base URL for the Databricks environment (dev, staging, prod).
Warehouse ID The specific SQL warehouse identifier for execution.
Catalog/Schema Scope The enforced namespace for data access.
Target Table Set The specific tables included in the query scope.
Output Defaults Includes objectives, assumptions, SQL drafts, and risk notes.

Databricks SQL Read-Only Advanced Features

  • Fail-closed allowlist enforcement for specific catalogs and schemas to prevent unauthorized data access.
  • Automated blocking of all mutating SQL keywords to ensure a purely read-only experience.
  • Detailed planning output for unsupported workflows like Jobs API or Unity Catalog lineage requests.
  • Robust risk assessment and rollback signals provided in every execution summary within Openclaw Skills.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*