Replace flag-file + heartbeat approach with the production dispatcher
pattern: poller triages notifications and spawns isolated agents
directly via openclaw cron. Adds assignment scan for self-created
issues. Response time ~15-60s instead of ~30 min.
- Rename 'PII Output Routing' → 'Sensitive Output Routing' throughout
- Expand scope to include secrets, credentials, API keys, flight numbers,
locations, travel plans, medical info
- Replace hardcoded 'Emails' heartbeat check with conditional language
('Notifications — whatever inbox sources you've integrated')
- Remove 'email' from heartbeat-state.json example
- Update cross-references in SETUP_CHECKLIST.md
- Replaced wake-event poller with flag-file approach (prevents DM spam)
- Added Option A (webhooks for VPS) vs Option B (poller for NAT)
- Documented the wake-event failure mode and why we switched
Humans also have limited state and drop things under cognitive load.
Universal make targets reduce context-switching cost across repos.
Agents can implement/enforce, but everyone benefits.
