Oracle Publishes Codex Workflow for Regional Analytics Workbooks

Oracle’s walkthrough shows how natural-language instructions can handle repeated report setup across regional audiences, while access and catalog changes remain governed actions.

By 2 min read
Oracle Publishes Codex Workflow for Regional Analytics Workbooks
Oracle Publishes Codex Workflow for Regional Analytics Workbooks

Listen to this story

The audio brief

About 1:40
0:001:40
Read transcript
Oracle has published a demonstration in which OpenAI Codex turns one Oracle Analytics Cloud workbook into three managed regional copies. The administrator starts with a completed workbook whose Region filter includes AMERICAS, APAC, and EMEA. Codex then creates a filtered copy for each region, names the copies, creates matching shared folders, and places each workbook in the right folder. The workflow adds a permissions step: each folder receives its corresponding regional application role, so users in a geography can access relevant reports without seeing data intended for other regions. That makes the example more than simple file duplication. It shows natural-language instructions coordinating workbook edits, folder setup, catalog organization, and access control in one repeatable process. But the automation does not bypass governance. Oracle’s Analytics MCP Server can save content, create, copy, move, or delete catalog items, and update access-control lists. Those catalog-changing operations are marked Experimental and require user approval, or an explicit approval flow. The server was introduced as a Preview capability in Oracle’s July update, and its actions inherit the authenticated user’s existing permissions. Oracle recommends checking both the workbook filters and the resulting folder and role permissions after a run. Codex is the client shown, but the workflow also supports Claude Desktop, Cline, and Copilot. The key constraint is clear: the setup can be automated, but administrative validation remains part of the workflow.

Story brief

3 key points

Oracle’s demonstration shows OpenAI Codex coordinating a multi-step Oracle Analytics Cloud administration workflow: generating three region-filtered workbook copies, creating matching shared folders, and applying AMERICAS, APAC, and EMEA application roles. The MCP Server can perform catalog and ACL changes, but those operations remain Experimental and require user approval or an explicit approval flow. Because...

  1. 01

    One source workbook becomes three regional copies, each with its corresponding filter and shared folder.

  2. 02

    Folder creation, copying, moving, deletion, saving, and ACL updates are marked Experimental catalog operations.

  3. 03

    Oracle introduced the Analytics MCP Server as a Preview capability in its July update.

Oracle has published a demonstration showing OpenAI Codex creating region-specific copies of an Oracle Analytics workbook, placing them in shared folders and assigning matching access roles. The example turns a repetitive administrator job into an AI-directed workflow.

The catalyst is familiar to teams that distribute the same analysis to different groups. An administrator may need to duplicate a workbook, adjust filters, create folders, save each version and configure access repeatedly for regions, business units or teams. Oracle’s example starts with a completed workbook whose Region filter includes AMERICAS, APAC and EMEA.

From one workbook to three managed copies

In the walkthrough, the administrator asks Codex to duplicate that workbook for every available region, name the copies by region and place each in a newly created shared folder bearing the same name. The resulting folders each contain a workbook with its corresponding regional filter applied.

Oracle then adds an access-control step. The prompt assigns AMERICAS, APAC and EMEA custom application roles to folders with matching names. Oracle describes the roles as a way to let users in each geography view relevant reports without seeing data intended for the other regions.

Oracle Analytics guidance showing that administrators should validate automation changes to catalog structures and permissions.
Oracle advises administrators to validate catalog and permission changes made through automation. Source: blogs.oracle.com.

Approval remains part of the workflow

The MCP server exposes operations that alter the Oracle Analytics Cloud catalog: saving content, creating folders, copying, moving or deleting items, and updating access-control lists, or ACLs. Oracle’s tool reference marks the catalog-modifying tools Experimental and requires user approval or an explicit approval flow for them.

Oracle introduced the MCP Server as a Preview capability in its July update. Its actions are constrained by the permissions assigned to the authenticated Oracle Analytics Cloud user, so connecting an AI client does not add authority beyond that user’s existing access.

The next check is administrative

Oracle recommends that administrators review and validate automated work, particularly when it changes catalog structure or permissions. A workbook can carry the intended filter while a folder or role assignment still needs administrative confirmation.

Codex is the client used in the walkthrough, but Oracle says the Analytics MCP workflow also supports Claude Desktop, Cline and Copilot. The example is therefore a reusable pattern for reducing repeated setup work.

Sources

  1. docs.oracle.comdocs.oracle.com
  2. blogs.oracle.comOracle Analytics July 2026 Update
  3. blogs.oracle.comAutomating Oracle Analytics Workbook Creation and Organization with MCP

Loading discussion...