diff --git a/work-items/open/20260527-194421-pod-orchestration-system-guidance/item.md b/work-items/open/20260527-194421-pod-orchestration-system-guidance/item.md index 6f74a320..f9658408 100644 --- a/work-items/open/20260527-194421-pod-orchestration-system-guidance/item.md +++ b/work-items/open/20260527-194421-pod-orchestration-system-guidance/item.md @@ -7,7 +7,7 @@ kind: feature priority: P2 labels: [pod, workflow, prompt] created_at: 2026-05-27T19:44:21Z -updated_at: 2026-05-27T19:44:43Z +updated_at: 2026-06-01T00:57:03Z assignee: null legacy_ticket: null --- diff --git a/work-items/open/20260527-194421-pod-orchestration-system-guidance/thread.md b/work-items/open/20260527-194421-pod-orchestration-system-guidance/thread.md index 5a835c3a..c1403734 100644 --- a/work-items/open/20260527-194421-pod-orchestration-system-guidance/thread.md +++ b/work-items/open/20260527-194421-pod-orchestration-system-guidance/thread.md @@ -55,4 +55,20 @@ AGENTS.md や workflow に multi-agent の運用は書かれているが、こ - TUI notification UI の変更。 +--- + + + +## Comment + +## Supplemental guidance from dogfooding + +Add two explicit rules to the Pod orchestration/system guidance: + +- A spawned Pod completion notification is delivered as a normal background signal. The parent does not need to keep the turn open or call tools solely to wait for it; it is acceptable to finish the current turn and handle the notification at the next natural point. +- Do not use `sleep`/polling loops just to wait for a Pod's output. If there is no other useful immediate work, return control to the user instead of blocking the turn; when the notification arrives, read the Pod output then. + +Rationale: during multi-agent work, waiting with `sleep` wastes the active turn and fights the notification model. The desired behavior is notification-driven follow-up, not artificial polling. + + ---