docs(agent): add concrete subagent monitoring thresholds

MEMORY.md

@@ -42,6 +42,7 @@
 - For non-trivial implementation work, treat `WIP.md` as the canonical state file and update it after each completed task/sub-task with status, concrete evidence, and the next recommended action.
 - If a subagent model choice causes execution/auth issues, prefer retrying implementation work on Codex GPT-5.4.
 - If a fresh implementation subagent stops making crisp progress, inspect once; if it is looping, not updating `WIP.md`, or returns an unusable result, kill it, verify the workspace directly, and finish the pass in the main session.
+- Monitoring cadence for fresh implementation subagents: first routine check at ~5 minutes if still running, inspect history at ~10 minutes, treat ~12/15 minutes as the suspicious/intervene threshold for narrow passes and ~20/25 minutes for medium bounded passes unless recent inspection shows crisp progress.
 
 ## Infrastructure notes worth remembering
 - Full `~/.openclaw` backups upload to MinIO bucket `zap` and are scheduled via OS cron every 6 hours.