HEARTBEAT.md - Accountability & Mission Adherence

Rotating Accountability Checks (2-3 per heartbeat, rotate through):

Charlie's Scope Protection

• Check recent meetings/emails: Is finance/ops routing to Charlie?
• If yes: Flag immediately, redirect to Vania
• Track violations in metrics/charlie-scope-violations.md

Team Capacity & Role Alignment

• Tin: Is she being overloaded? (already at capacity)
• Steve: Is he training or buried in admin?
• Kristin: Deepening relationships or drowning in systems?
• Malachi: Being rushed on Code 5? (respect his process)

Escape Velocity Indicators (check 1-2):

• Can Jedi Council make decisions without Quan's direct input?
• Are routine operations running without human intervention?
• Is Charlie designing again (not just managing)?
• Is Steve training without admin burden?
• Is Kristin relationship-building without system overhead?
• Is Quan architecting future vs. managing present?

Loss Function Alignment (sample recent decisions):

• Are decisions respecting priority order? (vitality → relational → sovereignty → business)
• Any skiing/movement opportunities being sacrificed for business?
• Is deep-work time protected? (one uninterrupted block per day)

External Proactive Checks (rotate, 2-4x per day):

• Email: Any urgent unread messages?
• Calendar: Upcoming events in next 24-48h?
• Weather: Relevant for skiing/travel decisions?

System Health (adaptive frequency - consolidate into heartbeat):

• Auto-commit status (hourly cron)
• Gmail OAuth tokens (3 accounts, hourly refresh)
• Zoho CRM tokens (when authorized, 50-min refresh)
• Email triage worker (quan@ztag.com processing)
• Cron jobs status (verify all scheduled tasks running)

Reporting frequency (adaptive):

• Days 1-7: Report every heartbeat (establish baseline)
• Days 8-30: Report daily (if 7+ consecutive successes)
• Days 31+: Report weekly (if 30+ consecutive successes)
• Always report immediately: Any failures, errors, or anomalies

Track in: memory/system-health-tracking.json

{
  "auto_commit": {"last_success": "timestamp", "consecutive_successes": 50, "report_frequency": "weekly"},
  "gmail_tokens": {"last_success": "timestamp", "consecutive_successes": 120, "report_frequency": "weekly"},
  "zoho_tokens": {"last_success": null, "consecutive_successes": 0, "report_frequency": "every"}
}

Background Work (do without asking):

• Review recent memory files, update MEMORY.md if significant
• Check git status, commit workspace changes
• Organize documentation if needed
• Review and integrate sub-agent findings
• Update system health tracking (success counters)

Tracking State

Track last check times in memory/heartbeat-state.json to avoid redundant work.
Track system health reliability in memory/system-health-tracking.json for adaptive reporting.

When to Reach Out vs. HEARTBEAT_OK

Reach out if:

• Charlie scope violation detected
• Team member overloaded beyond capacity
• Important email/calendar event
• Escape velocity regression (metric declining)
• Something interesting found that adds value

HEARTBEAT_OK if:

• Late night (23:00-08:00 PT) unless urgent
• Nothing new since last check (<30 min ago)
• Human clearly busy
• All checks green, no action needed

Quiet Hours

Respect: 23:00-08:00 PT (unless urgent)