← sources

What is the best way to handle independent agent communication?

source: reddit (reddit:ClaudeCode)published 2026-11-01 22:36ingested 2026-11-02 02:41

retained text

Let's say I have an extremely broad and high context project, and 10 remote machines are available to me for testing. I want one "manager" agent to be able to track what's going on with 10 different, independent agent sessions. I have a GUI that let's me check in on the progress of all 10 of them, and I want the 10 agent sessions to report back to the manager.

Is there a standardized best practice to facilitate this? The solution I have now is that the manager agent is listening for changes to text files in a folder that tracks changelogs for each of the 10 agents, but it feels not optimal.

document fetched 2026-11-02 00:34 · canonical https://reddit.com/r/ClaudeCode/comments/1xlgl7n/what_is_the_best_way_to_handle_independent_agent