Code Overlord vs Orca: agents, mobile and scheduled work

One instruction to Claude: have Codex implement the change, ask another agent to review it, correct the findings, run independent checks and get a final review. Setup, approvals and waits are shortened in this recording.

Watch the sidebar. Codex appears as a new session. The hierarchy shows it beneath the agent that called it. This is the coordinator creating its own worker.

I open Codex inside the same window. The implementation is ready, with thirty tests passing. The coordinator checks its result and moves the workflow forward.

Now a separate Claude reviewer joins the workspace. The coordinator sends it the review task. I switch back to the conversation, where both Claude sessions have their own tabs.

The review finds real issues in the keyboard handling and tests. Those findings go back to the original implementer through the coordinator.

Back in Codex, the requested corrections are complete and forty tests pass. The task stayed with its implementer; I did not copy the review between agents.

The coordinator creates another session for independent checks. Watch its tab appear alongside the reviewer. Each stage has its own agent and its own visible work.

The coordinator also created an independent checker. Here is its own session, confirming the test results and that the work remains uncommitted.

Back at the coordinator, here is Code Overlord’s completion message from the checker. The coordinator reads the result and sends the work into final review.

Final review catches one more keyboard-layout issue. The coordinator sends a precise correction back to Codex and keeps the reviewer ready to verify it.

Codex finishes that correction with forty-two passing tests. The completion returns to the coordinator, which starts a fresh check alongside final verification.

The reviewer confirms that all eight findings are fixed, forty-two tests pass, and the regression checks catch the broken versions. Its review is complete.

The fresh checker independently verifies the updated keyboard behavior and reports all five checks passed. Both the checker and reviewer have now finished.

The original coordinator brings the whole run together: implementation, review, corrections, checks and the final verdict. The changes are left here for my review.

Here is the completed team in its native hierarchy. One workflow instruction; real agents handling the handoffs. Everything stayed inside Code Overlord.

This is a real screenshot saved on my phone. I will send it straight to Claude in Code Overlord. Setup, approvals and waits are shortened.

Inside the workspace on my phone, I open the selected agent’s message box, choose an image from the gallery and add the screenshot.

I add a question and press Send. The attachment is going into this agent’s conversation.

The image arrives with my question in the selected desktop session. The workspace and other agents stay in view.

Claude reads the image: Dock Seven needs review, with twenty-seven parcels waiting. That is the detail visible in my screenshot.

Claude creates a new Codex session. Its native hierarchy shows who called it. I can open either conversation here.

From the phone, I ask Codex to open the original image and confirm. Claude passes that instruction and the image’s location to its worker.

Here is Codex’s answer in its own session: the same dock, both counts and the browser address. The second agent reports matching details.

I also ask Claude to keep a copy of the screenshot in the workspace. That gives me a file to return to.

I hover the returned file and choose View Scroll. There is the original phone screenshot, open inside Code Overlord.

Create a Ritual and choose its agent. This demonstration uses two fictional feedback reports. One is approved for development; the other stays in triage.

Tell the Ritual to preserve each report, attach a requirement, and develop approved work through an implementer and a separate reviewer. Save the note references so later runs can continue the same work.

Save the Ritual. Its first run is due now. The saved instruction belongs to the schedule, ready for the agent that runs it.

For this short demonstration, change the repeat interval from thirty minutes to one minute. Choose a schedule that suits your real workflow.

Open the saved Ritual to see its scheduled agent at work. It has the instruction already and checks the feedback before handing out any development.

Two notes appear in the notepad beside the conversation. Each preserves the original feedback and has an attached requirement. Only the search item is approved for development.

The scheduled agent opens an implementer. Watch the new tab appear inside Code Overlord. Its assignment includes the approved requirement and the exact project folder.

The implementation finishes and its completion message returns to the scheduled agent. A separate reviewer then appears, with instructions to inspect the actual changes and run its own checks.

The agent collects both results: twenty-three tests pass, and the independent review reports no actionable findings. It writes the checked outcome back to the original feedback item.

Open the attached requirement. The original report, approval and acceptance checks are readable beside the conversation, with the completed outcome still in the notepad.

The other requirement remains in triage. Recording feedback and authorising development are separate steps in this saved workflow.

A fresh scheduled session checks the same notes, attachments and completed result. It launches no new development workers and creates no duplicate feedback items.

The demonstration Ritual is now disabled. Its notes and reviewed result remain available. One saved workflow connected the schedule, the agents and the original feedback.