Update spawn-issues to event-driven pattern

Replace polling loop with task-notification based orchestration.
Background tasks send notifications when complete - no need to poll.

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

commands/spawn-issues.md

@@ -137,22 +137,24 @@ Process:
 Work autonomously. If blocked, note it in PR description and report status as partial/failed.
 ```
 
-### Step 4: Poll and Pipeline
+### Step 4: Event-Driven Pipeline
 
-Run a polling loop until all issues reach a terminal state (approved, failed, or max iterations):
+**Do NOT poll.** Wait for `<task-notification>` messages that arrive automatically when background tasks complete.
+
+When a notification arrives:
+1. Read the output file to get the result
+2. Parse the result and print status update
+3. Spawn the next stage (reviewer/fixer) in background
+4. Continue waiting for more notifications
 
 ```
-while any issue not in terminal state:
+On <task-notification> for task_id X:
-  for each issue:
+  - Find which issue this task belongs to
-    check TaskOutput (block=false) for current task
+  - Read output file, parse result
-
+  - Print status update
-    if task completed:
+  - If not terminal state, spawn next agent in background
-      parse result
+  - Update issue state
-      print status update
+  - If all issues terminal, print final summary
-      start next step (spawn next agent)
-      update issue state
-
-  brief pause before next poll iteration
 ```
 
 **State transitions:**
@@ -164,7 +166,7 @@ implementing → (worker done) → reviewing → (approved) → DONE
            → (worker failed) → FAILED
 ```
 
-**On each completion, print status:**
+**On each notification, print status:**
 
 ```
 [#42] Worker completed → PR #55 created, starting review