Security recommendations for Agent Studio

Secure the Agent Studio security framework by deploying the containerized ML runtime image and configuring strict project access controls.

Migration and deployment

Runtime mode only (recommended) – The use of the prebuilt, containerized ML runtime image is the recommended method for deploying Agent Studio. All new features and security improvements is available exclusively through this method. The previous methods of deployment (from source code) and AMP mode are deprecated.
Use service accounts for external services – If your shared Agent Studio instance connects to external services like Cloudera Data Warehouse or Datalake using JWT tokens or Kerberos credentials, deploy the workflow using a service account.
note
For security and auditing reasons, operating in a shared environment without a service account is not recommended. In this scenario, access is granted using the credentials of the application creator rather than your credentials.
Do not bypass application authentication – Agent Studio is designed to rely on the Cloudera AI Workbench for authentication. Bypassing this authentication mechanism during Agent Studio application deployment is not recommended.

Limit collaborators (RBAC Gap) – Agent Studio currently lacks a native Role-Based Access Control (RBAC) mechanism. To manage access effectively, you must create private projects and strictly limit the number of collaborators to those who have access to the project.
Minimal permissions for custom tools – When developing and configuring custom tools, such as MCP servers, adhere to the principle of least privilege. Custom tools run in a sandboxed environment with limited default access. Provide only the minimal necessary permissions to the custom tools.