Repository navigation
Replies: 1 comment 2 replies
|
Hii @jwade83, Your architecture is quite close to what Codex already supports. I would separate the responsibilities between the Codex runtime and your external supervisor.
One important distinction: resuming a Codex thread is supported, but automatically routing results into an arbitrary desktop conversation requires using the same intended thread and compatible App Server environment. I would validate this integration boundary before building additional infrastructure around it. Useful references:
I would recommend starting with the App Server and Python SDK rather than building another agent runtime around Codex. Your existing NATS/JetStream layer can complement the harness instead of duplicating its responsibilities. If this answer solves your doubts, could you please mark it as resolved? If the issue still persists, feel free to share the updated error so we can investigate further. |
Uh oh!
There was an error while loading. Please reload this page.
We’re building a personal coordination system around existing Codex workers, a Linux NATS/JetStream hub and durable project records. Bounded tests have recovered activities by name and returned results automatically to a desktop chat.
Inspired by Harness engineering, we want an agent to retain an objective, discover necessary tasks, execute and review them, then choose the next useful action across runs.
Which responsibilities can the supported Codex SDK/app-server interfaces handle directly, and which require an external supervisor? In particular, what supported interface should connect durable external events to resumed work and delivery into the intended desktop conversation?
We would appreciate a reference implementation or a maintainer’s review of a small architecture brief. Our priority is reusing the existing harness and understanding its actual integration boundaries.
All reactions