databricks-genie-agents
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the official Databricks CLI and the
jqutility to perform workspace management tasks, such as creating, exporting, and querying Genie Agents. These operations are performed within the user's authenticated environment using standard administrative tools. - [EXTERNAL_DOWNLOADS]: The skill's documentation directs the user to install official Databricks libraries (
databricks-openai,databricks-sdk) from the standard Python package registry. These resources are provided by the skill's author and are standard for the platform. - [DYNAMIC_EXECUTION]: Python helper scripts are provided in the references to assist in generating the complex
serialized_spaceJSON payloads. These scripts run locally and use standard libraries (json,uuid) to facilitate configuration management. - [INDIRECT_PROMPT_INJECTION]: As the skill facilitates agents that translate natural language to SQL, it proactively mitigates injection risks by mandating "Hard Rules" that restrict the agent to read-only SQL operations (e.g.,
SELECT,DESCRIBE). It also establishes human-in-the-loop approval gates for all live agent updates to ensure user oversight.
Audit Metadata