Primary Objective

Execute strategic oversight and final decision-making authority for all active UMS and VGP systems during the nightly window (10:00 PM – 6:00 AM CT). Responsibilities include system health verification, management accountability review, project governance, key approvals, and shift handoff to the morning management team.

Business Function Protected

• Strategic continuity and leadership presence during night hours
• System health of all live platforms (ERP, dashboards, automations)
• Management accountability — ensuring Ernesto, Myrille, and Reymark complete their assigned functions
• Final decision authority for approvals that cannot be delegated
• Company-wide KPI tracking and course corrections
• Project governance — no deployment or structural change proceeds without EJ sign-off

Revenue Impact

• An unmonitored ERP overnight risks data integrity failures, missed fraud alerts, and uncaught financial anomalies.
• Delayed management decisions cost compounding time on every high-priority item — every 24-hour deferral multiplies the backlog.
• Without a nightly system health check, issues discovered at opening can take 2–4 hours to resolve, directly impacting morning operations and customer commitments.
• An unclear morning handoff creates duplicate effort and confusion across all three managers at shift start.

What Happens If This Role Fails

• Unapproved decisions stall the following day's operations.
• System failures or data errors go undetected until morning — compounding by 6–8 hours.
• Management accountability slips — no one is holding the team to their commitments.
• ERP, dashboards, and automations run without strategic oversight.
• Handoff to morning team is unclear, causing duplicate effort and confusion.
• Project governance breaks down — staging features go live without approval.

Health-Adjusted Operations

This Role IS Responsible For

• Final approval authority on all items requiring owner-level sign-off
• Strategic oversight and direction for all active UMS and VGP systems
• Monitoring that Ernesto, Myrille, and Reymark are executing their respective functions
• Reviewing and verifying system health across all live platforms during the nightly window
• Issuing directives — every observation EJ makes becomes a directive, not a personal task
• Documenting all major decisions made during the shift in the Decision Log
• Delivering a shift handoff note to morning management before 6:00 AM
• Reviewing and approving new project milestones, live deployments, and go/no-go decisions
• Reviewing the AI Transformation Tracker to ensure automation progress is on track
• Reviewing the Bonus Tracker for alignment with actual performance metrics

This Role is NOT Responsible For

• Personally resolving operational issues — every issue gets assigned to the responsible manager
• Handling day-to-day tasks that belong to Ernesto (ERP/systems), Myrille (KPIs), or Reymark (VGP/HR)
• Direct customer or vendor interactions
• Data entry, lead qualification, or expense reporting
• Technical execution (coding, deployments, builds) — assign to Ernesto or AI Bot
• Creating content, SOPs, or documents directly — assign to AI Bot or relevant manager
• Running or modifying automations without Ernesto's technical confirmation

Departmental Boundaries

• EJ → Ernesto: ERP decisions, system architecture, technical approvals, live deployments
• EJ → Myrille: Company KPI review, anti-regression checks, growth direction
• EJ → Reymark: VGP operations, HR matters, SOP rollout oversight
• EJ → Austin: Structural decisions, ownership-level escalations, major financial approvals
• EJ → AI Bot: Report generation, SOP builds, data analysis, memory queries, dashboard refreshes
• EJ → All Managers: Every issue observed becomes a written directive — documented and assigned, never personally completed

Output 1: Nightly Decision Log

• Recipient: All active managers (Ernesto, Myrille, Reymark) + on-file record
• Format: Bulleted list — what was decided, who owns it, and the expected next action
• Deadline: Compiled before 5:30 AM
• Quality Standard: Every decision includes: outcome, owner, and next-action deadline. No mental notes.
• Verification: Logged in Management GC or workspace memory file
• System Logged In: Management GC + workspace memory/YYYY-MM-DD.md

Output 2: Management Accountability Review

• Recipient: Ernesto, Myrille, Reymark
• Format: Summary of what each manager completed vs. expected; any correction directives issued
• Deadline: Completed by 11:30 PM (first 90 min of shift)
• Quality Standard: Every stalled item has an assigned manager, correction directive, and deadline
• Verification: Each item has a written directive in the Management GC
• System Logged In: Management GC

Output 3: Morning Handoff Note

• Recipient: Morning management team (Ernesto, Myrille, Reymark)
• Format: "EJ EON — [date] — Top 3 priorities: [1] [2] [3] — Flags: [any] — Signing off [time]"
• Deadline: Sent before 6:00 AM — NON-NEGOTIABLE
• Quality Standard: Clear, actionable, no ambiguity about what needs to happen first each morning
• Verification: Message visible in Management GC or Telegram before 6:00 AM
• System Logged In: Management GC / Telegram

Output 4: System Health Verification

• Recipient: Self + management team (if issues found)
• Format: Status check across all live platforms — flagged issues assigned to responsible manager
• Deadline: Completed by 10:45 PM (first 45 min of shift)
• Quality Standard: No RED system flag left unassigned at shift end
• Verification: All dashboards loading, no stale data, no unresolved alerts
• System Logged In: ERP Ops Console + all dashboards

Input 1: Management EOD Summaries

• Source: Ernesto, Myrille, Reymark — end-of-day completion reports
• Required Format: What was completed, what is pending, any blockers
• Valid If: Submitted before or during EJ's shift window
• Invalid If: Missing, blank, or contains unresolved urgent items without documentation
• If Missing: Request immediately via Management GC at shift start. Flag in shift Decision Log.
• If Incorrect: Escalate directly to the relevant manager for clarification. Do not assume completion.

Input 2: System Alerts & Automation Reports

• Source: UMS ERP (fraud scan, health monitor), AI Bot, Cloudflare Pages
• Required Format: Alert message + source system + timestamp
• Valid If: Timestamped and clearly sourced from a live system
• If Missing: Manually verify via ERP Ops Console and Ops Dashboard during startup
• If Incorrect: Cross-reference with ERP API directly; assign Ernesto to investigate

Input 3: Strategic Directives (from Austin or Ownership)

• Source: Austin Hamil / George — via Management GC
• Required Format: Written directive in GC — verbal directives do not count until confirmed in writing
• If Missing: Proceed with standard SOP priorities until directive is received
• If Unclear: Request clarification via Management GC before acting

Input 4: Project Status Updates

• Source: Ernesto (ERP/systems), Myrille (KPIs), Reymark (VGP operations)
• Required Format: Per active project — status, blockers, next step
• Valid If: Covers all active projects under EJ governance
• If Missing: Request update during accountability review block (10:45 PM)

System Failure Protocols

• ERP Down: Message Ernesto immediately regardless of hour. Use Ops Dashboard + Owner Dashboard for high-level status. Defer ERP-specific approvals until restored.
• Dashboard Data Stale: Assign Ernesto to run the refresh script for affected dashboard. Pull from Google Sheets source if urgent.
• AI Bot Unavailable: Complete review manually using direct dashboard URLs. Flag for Ernesto to check OpenClaw gateway status.
• Automation Alert Missing: Manually check the relevant ERP section. Assign Ernesto to verify cron job status.

Time Required: 15 minutes. All steps must be completed before any task block begins.

1. Open UMS ERP (ums-erp.pages.dev) — navigate to #mission (Ops Console)
2. Verify all department live data indicators are active — no "NO LIVE DATA YET" panels
3. Open Owner Independence Dashboard (ums-owner-dashboard.pages.dev) — confirm data is current
4. Open Ops Dashboard (ums-ops-dashboard.pages.dev) — note Company Health hero score
5. Open Management Command Center (ums-command-center.pages.dev) — check KPI and Triage tabs
6. Review all unread messages in Management GC — note any unresolved items from day shift
7. Check for any new fraud alerts — navigate to ERP Analytics → Financial Controls
8. Check AI Transformation Tracker (ums-ai-tracker.pages.dev) — any automation failures overnight?
9. Confirm EOD summaries received from Ernesto, Myrille, and Reymark — all three required
10. Identify all items requiring EJ decision tonight — write them as tonight's decision queue
11. Confirm today's top 3 priorities from the previous morning's handoff note
12. Send to Management GC: "EJ online — [HH:MM PM] — reviewing systems"

Task 1: Nightly Startup (10:00–10:15 PM)

See Section 6 for full checklist. Required systems: all dashboards, ERP, Management GC, AI Bot.

Task 2: System & Dashboard Review (10:15–10:45 PM)

ERP Review Steps

1. Navigate to ums-erp.pages.dev/#mission — review all active department panels
2. ERP Analytics → Financial Controls — review AR aging, AP aging, cash position, margin alerts
3. ERP Analytics → Task Queue — any overdue tasks with no completion verified?
4. Check ERP Approvals queue — any pending approvals requiring EJ sign-off?
5. Review Import dept — open PO status, any OUTSTANDING items past expected delivery?
6. Review Sales dept — any invoices past 48 hours without fulfillment action?
7. Review Dispatch dept — any unreconciled pickups from today?
8. Check Fraud Detection — any HIGH-severity findings that triggered an alert tonight?
9. Document all flags: "[Dept] — [Issue] — [Severity] — [Assigned To]"

Dashboard Review Steps

1. Ops Dashboard — record Company Health hero score
2. Check each dept card: Dispatch, Import, Inventory, Warehouse, Marketing — flag any RED or STALE
3. Owner Dashboard — confirm inventory and sales figures match expectations
4. Command Center — review KPI tab for any metrics in WARNING or CRITICAL
5. Analytics Command (EJO) — review for any data anomalies or unusual trends

Task 3: Management Accountability Review (10:45–11:30 PM)

Ernesto (ERP & Systems Manager)

1. Did Ernesto complete all ERP tasks assigned for today?
2. Are all staging → live promotions documented with clear go/no-go criteria?
3. Are all system health monitors running (fraud scan cron 62c14ae0, health monitor a8b5411d)?
4. Any ERP bugs reported today that need EJ prioritization or sign-off?
5. Issue correction directive for any incomplete item: "Ernesto — [task] — expected by [deadline]"

Myrille (Company KPIs & Anti-Regression)

1. Did Myrille complete KPI review and company health check today?
2. Were any anti-regression issues flagged? What is the current resolution status?
3. Is the VA Performance Hub (ums-commitments-tracker.pages.dev) data current?
4. Any growth tracking or reporting items pending EJ review or decision?
5. Issue correction directive for any incomplete item: "Myrille — [task] — expected by [deadline]"

Reymark (VGP Operations & HR)

1. Did Reymark complete VGP operations review and HR task follow-ups?
2. Are all HR items (timekeeping, employee issues, compliance) handled?
3. Is the individual SOP rollout on track — have role cards been distributed and acknowledged?
4. Any VGP-specific flags or escalations requiring EJ attention or decision?
5. Issue correction directive for any incomplete item: "Reymark — [task] — expected by [deadline]"

Task 4: Active Projects Review (11:30 PM–12:30 AM)

For each active project, confirm: current status (ON TRACK / STALLED / NEEDS DECISION), any blocker requiring EJ decision, and any action items assigned.

Active Projects Checklist

1. UMS ERP (Live): Open bugs? Pending feature decisions? Deployments awaiting EJ sign-off?
2. Price Intel System (Staging): Awaiting EJ go/no-go for live promotion — make or formally defer the call tonight
3. Warehouse UI Concept B (Staging): Awaiting EJ approval — review and decide or defer with reason
4. Individual SOP Rollout: Is each team member's SOP current and acknowledged by the role holder?
5. Bonus Tracker: Any disputes, discrepancies, or upcoming payout deadlines requiring review?
6. VA Performance Hub: Any accountability flags needing escalation to Austin?
7. AI Transformation Tracker: Is the progress log current? Any automations stalled?
8. ERP SOP Manager: Any SOPs needing update after recent ERP changes (SOP_BACKLOG.md)?

Task 5: Decisions & Approvals Block (12:30–1:30 AM)

1. Work through the decision queue identified during startup
2. For each item requiring approval — apply the 9-Field Approval Template (see Section 12)
3. Approve, deny, or defer each item with a documented reason
4. Send all approval decisions to the requesting manager via Management GC
5. For items needing ownership-level review from Austin — draft and send the escalation now
6. Document all decisions in tonight's Decision Log before moving to the next block

Decision Priority Order

• P1 — Resolve before 2:00 AM: Anything blocking a manager from completing their shift tasks
• P2 — Resolve before 3:00 AM: Revenue-impacting decisions (customer approvals, pricing, fulfillment)
• P3 — Resolve same shift: System deployment approvals (ERP features, staging promotions)
• P4 — Document and schedule: Strategic reviews (project direction, team structure, new initiatives)

Task 6: Strategic Review & Direction (1:30–3:00 AM)

1. Review company trajectory — are we on track for current month targets across all departments?
2. Review AI Tracker — which automations are yielding results? Which are stalled or failing?
3. Review staffing and headcount items — apply profit-per-employee benchmark ($3K–$5K/month mature)
4. Identify 1–3 strategic improvements with the highest impact this week — document as directives
5. Review Bonus Tracker — are bonuses aligned with verified performance metrics?
6. Review Individual SOP status — is the team operating from current, accurate SOPs?
7. Draft any new directives, policy changes, or system priorities arising from the strategic review

Task 7: Deep Work / Flexible Block (3:00–4:30 AM)

This block is health-dependent and flexible. Use it for:

• Reviewing detailed reports or audit outputs that require focused reading
• Reading and responding to complex management documents or SOP drafts
• Making deferred decisions that require more thought or research
• Reviewing financial statements, procurement reports, or performance data
• Any deep strategic work assigned to EJ that cannot be delegated

Task 8: Shift Close Prep (4:30–5:30 AM)

1. Review all decisions made tonight — confirm each is documented in the Decision Log
2. Review all directives issued tonight — confirm each has a named owner and deadline
3. Review all escalations sent — confirm receipt acknowledged or follow-up is scheduled
4. Confirm no system is in RED status at shift end — assign morning follow-up if RED remains
5. Set top 3 priorities for the morning management team
6. Draft morning handoff note using the Output 3 format from Section 3
7. Verify all dashboards are loading correctly and data appears current

Task 9: Final Scan & Shift End (5:30–6:00 AM)

1. Final pass through: ERP Ops Console, Ops Dashboard, Owner Dashboard
2. Confirm no overnight alerts, system failures, or urgent issues were missed
3. Send morning handoff note to Management GC: "EJ EON — [date] — [3 priorities] — signing off [time]"
4. If any critical unresolved item exists — contact Austin directly before signing off
5. Log shift summary in workspace memory/YYYY-MM-DD.md

Decision 1: System Alert — Investigate or Escalate

• IF: Any dashboard shows RED status OR ERP fraud scan shows HIGH finding
• THEN: Investigate source. If resolvable by reviewing data — document and assign to responsible manager with a deadline.
• ESCALATE IF: Issue requires immediate technical action (ERP down, data corruption, security breach) — contact Ernesto immediately regardless of hour.
• DEFAULT: Document in shift Decision Log, assign morning follow-up to responsible manager.

Decision 2: Manager Accountability — Correct or Escalate

• IF: Any of the three managers has not completed their assigned tasks for the day
• THEN: Issue a written correction directive with a specific deadline.
• ESCALATE IF: Same item is missing for 3+ consecutive shifts — initiate formal corrective review with Austin.
• DEFAULT: Issue directive. Never personally complete the task.
• RULE: Every observation EJ makes becomes a directive. EJ does not personally resolve what management should resolve.

Decision 3: Project Go/No-Go Decision

• IF: A project or feature is ready for live promotion (e.g., Price Intel, Warehouse UI Concept B)
• CRITERIA: Is it tested in staging? Is there a rollback plan? Is there a clear business need now?
• APPROVE: All three criteria met — document approval and notify Ernesto to deploy.
• DEFER: Any criterion not met — document the specific reason and set a re-review date.
• DEFAULT: Defer. Never approve a live deployment without verified testing and a rollback plan.

Decision 4: Hiring / Headcount Decision

• IF: A manager requests a new hire or additional staff
• CRITERIA: Apply profit-per-employee benchmark ($3,000–$5,000/month mature; $1,500+ early stage). Is headcount tied to a measurable bottleneck, revenue opportunity, or documented workload threshold?
• APPROVE: If all criteria met — document approval and escalate to Austin for final structural sign-off.
• DENY: If headcount growth would outpace current profit growth — document denial with reason.
• DEFAULT: Defer to Austin for all structural headcount decisions.

Decision 5: Escalation to Austin / Ownership

• IF: Any structural decision, major financial commitment, or ownership-level matter arises
• THEN: Document the item clearly (context, options, EJ's recommendation) and send to Austin via Management GC.
• IF NO RESPONSE WITHIN 24 HOURS: Re-escalate with "FOLLOW-UP — [item] — awaiting decision — [impact if delayed]"
• DEFAULT: Never make ownership-level decisions unilaterally. Document and escalate.

• GREEN: Continue normal operation. Log status in shift Decision Log.
• YELLOW (Warning): Note in Decision Log. Identify cause. Self-correct next shift. Notify Austin if pattern continues 2+ consecutive shifts.
• RED (Critical): Flag immediately. Notify Austin via Management GC same shift. Include cause and proposed corrective action.
• RED REPEATEDLY (3+ shifts): Escalate to Austin for formal corrective review. SOP gap analysis required.

KPI 6 — Project Governance Enforcement

• GREEN: All active projects reviewed and status logged in Decision Log.
• YELLOW: 1 project missed. Documented reason in shift log. Catch up next shift.
• RED: Project not reviewed 5+ shifts. Added to P1 decision queue. Austin notified.

KPI 7 — Health-Adjusted Completion Enforcement

• GREEN: Core 3 completed. Full shift when energy allows.
• YELLOW: Deep work skipped with documented note. Core 3 still complete.
• RED: Core 3 skipped without health reason. Austin notified. Catch-up shift scheduled.

Automation 1: Daily Fraud Scan

• Trigger: Cron job at 7:00 AM CT daily (automation ID: 62c14ae0)
• Condition: ERP financial data scanned for anomalies and fraud signals
• Action: HIGH-severity findings trigger Telegram alert to EJ and management team
• Assigned To: EJ reviews during startup checklist (step 7)
• Approval Required: No — scan runs automatically on schedule
• Escalation: HIGH finding detected → assign Ernesto to investigate same night
• Log Location: ERP Analytics → Financial Controls
• If Fails Silently: No alert received but scan was scheduled — EJ manually checks Financial Controls during startup. If no scan log visible — assign Ernesto to verify cron job 62c14ae0.

Automation 2: System Health Monitor

• Trigger: Every 30 minutes (automation ID: a8b5411d)
• Status: DISABLED — pull_sales.sh returned HTTP 400 errors on missing tab name
• Manual Coverage: EJ manually verifies system health during startup until Ernesto re-enables with graceful 400 fix
• Action Required: Ernesto — re-enable after fixing graceful HTTP 400 handling in pull script
• Escalation: If ERP behaves unexpectedly between manual checks — notify Ernesto immediately

Automation 3: ERP Data Backup & Recovery

• Trigger: Ernesto-managed Cloudflare D1 backup schedule
• Assigned To: Ernesto (manages and monitors)
• EJ Monitoring: If ERP data appears corrupted or missing — Ernesto executes recovery_drill.sh
• Escalation: Any ERP data integrity issue → Ernesto immediate contact, regardless of hour
• Recovery Script: outputs/ums-erp/recovery_drill.sh (non-destructive — validates D1 restore, API smoke test, failover sim)

Automation 4: Price Intel Weekly Scan

• Trigger: Weekly automation (ID: 5b0af746)
• Status: STAGING ONLY — awaiting EJ go/no-go approval for live promotion
• Action: Scans competitor prices, stores entries, compares vs UMS buying/selling prices
• EJ Decision: Review status during Active Projects block (Task 4) — approve, defer, or formally deny live promotion
• On Approval: EJ approves → Ernesto promotes Price Intel to live ERP

Automation 5: VA Performance Hub Refresh

• Trigger: bash outputs/commitments-tracker/refresh.sh (manual or scheduled)
• Action: Pulls Google Sheets data, rebuilds site, publishes to ums-commitments-tracker.pages.dev
• Assigned To: Ernesto (technical execution); Myrille (data oversight)
• EJ Monitoring: If VA Performance Hub looks stale during accountability review — assign refresh to Ernesto

Actions Requiring EJ Approval

• Any live deployment or production promotion of ERP features from staging
• Go/no-go decisions on staging-to-live transitions (Price Intel, Warehouse UI Concept B, etc.)
• New hire or headcount additions (apply profit-per-employee benchmark first)
• Changes to individual SOP structure or role scope
• Bonus payout approvals or bonus structure modifications
• New system, platform, or tool adoption (integrations, software, subscriptions)
• Formal corrective review initiation for any team member
• Major strategic pivots, project cancellations, or scope changes
• Any financial commitment outside normal operations

Approval Request Format — 9-Field Template (Required for All Approval Requests to EJ)

• ACTION_ID: ACTION_[MM-DD-YYYY]_[TYPE]
• Requested Action: [exact description of the action]
• Business Reason: [why this is necessary now]
• Risk Level: Low / Medium / High
• Amount: $[X] or N/A
• Customer/Dept: [who is impacted]
• Data Affected: [what records or systems are involved]
• Required Approver: EJ Espiritu
• Deadline: [time by which action must be taken]
• Approval Status: PENDING / APPROVED / DENIED

Actions Requiring Austin (Ownership) Approval

• Structural headcount decisions (new permanent roles or role eliminations)
• Major financial commitments or vendor contracts
• Ownership-level policy changes or company-wide structural decisions
• Any decision EJ is uncertain about — escalate rather than guess

Austin Hamil — Senior Management / Ownership

• You Receive: Structural directives, ownership-level approvals, strategic guidance
• You Send: Escalations, formal corrective review requests, major approval requests, critical system flags
• Channel: Management GC / Direct message
• Timing: P1 escalations — same shift. Non-urgent escalations — within 24 hours.
• If No Response to Escalation: Re-send after 24 hours: "FOLLOW-UP — [item] — awaiting decision — [impact if delayed]"

Ernesto — ERP & Systems Manager

• You Receive: ERP status updates, system flags, technical completion reports, bug reports
• You Send: ERP decisions, deployment approvals, technical directives, system priorities, go/no-go calls
• Channel: Management GC or direct Telegram
• Timing: Directives issued during accountability review (10:45 PM). Technical emergencies — any time, immediately.
• If No Response to Directive: Follow up once after 2 hours. Escalate to Austin if critical operations are blocked.
• Coordination: All ERP builds, deployments, SOP rebuilds, dashboard refreshes, and automation changes go through Ernesto or the AI Bot on EJ's instruction

Myrille — Company KPIs & Anti-Regression

• You Receive: Company health reports, KPI status, growth tracking updates, anti-regression flags
• You Send: KPI review outcomes, growth direction, anti-regression directives, accountability corrections
• Channel: Management GC
• Timing: Review completed during accountability block. Non-urgent corrections noted for morning team.
• If No Response: Follow up next shift. Escalate to Austin if same item is missed 3+ shifts.
• Coordination: Myrille is the signal layer — she flags when metrics fall below threshold; EJ decides the strategic response

Reymark — VGP Operations & HR

• You Receive: VGP status updates, HR issue reports, SOP distribution confirmations, attendance flags
• You Send: VGP directives, HR policy guidance, SOP rollout instructions, accountability corrections
• Channel: Management GC
• Timing: Review completed during accountability block. HR-critical items escalate to Austin as needed.
• Coordination: Reymark owns VGP operations and HR execution — EJ owns the governance layer above both

UMS Management AI Bot

• You Receive: Automated alerts, system reports, analysis outputs, SOP builds, dashboard refreshes
• You Send: Analysis requests, SOP build requests, report generation, memory queries, data pulls
• Channel: Direct Telegram / OpenClaw
• Usage Rule: If any task can be completed with data available to the AI — assign it to the bot before doing it manually. The bot handles SOP builds, report generation, dashboard refreshes, memory logging, and analysis.
• Timing: Available 24/7. Responds in real time during active sessions.

I Receive From

• Austin: Structural directives, ownership decisions — via Management GC, any time
• Ernesto: ERP system status, technical flags, deployment readiness reports — before or during shift
• Myrille: Company KPI status, anti-regression reports, company health score — before or during shift
• Reymark: VGP status, HR updates, SOP acknowledgment confirmations — before or during shift
• AI Bot: Automated alerts, analysis reports, system health summaries, SOP documents
• ERP System: Fraud alerts, approval queue items, financial control outputs, task queue status
• All Dashboards: System health data, KPI metrics, company health score, project status

I Produce For

• Ernesto: ERP decisions, deployment approvals, technical directives, system priorities
• Myrille: KPI review outcomes, growth directives, accountability corrections, anti-regression responses
• Reymark: VGP directives, HR guidance, SOP rollout instructions, accountability corrections
• Austin: Escalations, formal reviews, major approval requests, strategic recommendations
• Morning Team (all managers): Shift handoff note with top 3 priorities — before 6:00 AM every shift
• All Managers: Decision Log — all approvals, denials, and directives from the shift

Dependency Risks

• EOD Summaries Not Received: Accountability review cannot be completed accurately. Request summaries immediately at shift start. Flag in Decision Log if not received by 10:45 PM.
• System Alert Not Actioned: Issue compounds overnight unchecked. EJ must flag and assign within 30 minutes of detection — no exceptions.
• Handoff Note Not Sent: Morning team starts without clear priorities — the highest-impact failure mode. Always send the handoff note, even if abbreviated. One sentence is better than nothing.
• Decision Queue Not Cleared: P1 blockers delay the next day's operations across all departments. Resolve all P1 decisions before 2:00 AM.
• Manager Non-Responsive: Issue goes without resolution. Escalate to Austin after 2+ follow-ups with no response on a critical item.

ERP System Unavailable

• Detection: ums-erp.pages.dev fails to load or returns an error
• Immediate Action: Check Cloudflare status page. Message Ernesto immediately — regardless of hour.
• Backup: Use Ops Dashboard + Owner Dashboard for high-level status. Defer all ERP-specific approvals until restored.
• Escalation Deadline: Ernesto notified within 5 minutes of confirmed failure
• Log It: Document in Decision Log: system, time detected, who was notified

Dashboard Data Stale or Missing

• Detection: Any dashboard shows data from prior day or displays a "stale" indicator
• Immediate Action: Assign Ernesto to run the refresh script for the affected dashboard
• Backup: Pull metrics directly from the Google Sheets source if known and urgent
• Log It: Document which dashboard, when detected, and what action was assigned

Management Team Non-Responsive

• Detection: Manager does not respond to directive or request within expected timeframe
• Immediate Action: Follow up once directly (Management GC message or direct Telegram)
• Escalation: If critical item remains unresolved after 2 follow-ups — escalate to Austin with full context
• NEVER: Personally complete the task. Issue a stronger directive with deadline and state the consequence of non-completion.

Automation Failure Detected

• Detection: Expected alert or report not received, or cron log shows a failure
• Immediate Action: Assign Ernesto to diagnose the specific automation and its cron job
• Backup: Manually complete the check that the automation was performing during startup
• Log It: Document automation name, expected action, what was missing, and Ernesto's assignment

EJ Health Unavailability

• Detection: EJ is unable to complete the nightly shift due to health
• Immediate Action: Message Austin to cover as backup owner
• Minimum Before Signing Off: Send a brief status message to Management GC noting unavailability and any P1 items requiring immediate attention
• Recovery: Next available shift — review all missed items and issue catch-up directives. No guilt, no pressure — health comes first.

• How Error is Identified: Discovered by EJ during shift review, or flagged by a manager, the AI bot, or an automated alert
• Who Verifies: EJ confirms the error by checking the live system — not a report or screenshot alone
• Who Approves Correction: EJ for operational errors. Austin for structural or financial errors.
• How Correction is Made: Written directive issued to the responsible manager: what was wrong, what correction is required, and the deadline
• How Correction is Logged: Decision Log entry + Management GC message; relevant Google Sheet or ERP record updated by the responsible manager
• Who is Notified: The responsible manager always. Austin if the error affected revenue, data integrity, or compliance.
• Prevention: The accountability review block (Task 3) is the primary error-prevention mechanism. Completing this block consistently catches the majority of errors before they compound. Consistent completion of the startup checklist catches system failures before they impact the day.

Shift Completion Quality Gate

• Startup checklist completed (Section 6) — all 12 steps
• All three managers reviewed — Ernesto, Myrille, Reymark
• All system flags documented and assigned to a named owner with a deadline
• All P1 and P2 decisions resolved before shift close
• Morning handoff note sent before 6:00 AM
• Decision Log updated and complete — every decision, approval, and directive logged
• No RED system status left unassigned at shift end

Management Accountability Quality Gate

• Each manager's EOD summary received and reviewed
• Each incomplete item has a written directive with a named owner and specific deadline
• No mental notes — every directive is documented in writing in Management GC
• Directive format used: "[Manager] — [task] — expected by [deadline]"

Project Review Quality Gate

• Every active project has a current status: ON TRACK / STALLED / NEEDS DECISION
• Every NEEDS DECISION project received a decision or a documented deferral with reason and re-review date
• No project goes 5+ shifts without a status review

Approval Quality Gate

• All approval requests received a response: APPROVED / DENIED / DEFERRED WITH REASON
• All approvals used the 9-Field Template
• No live deployments approved without verified staging testing and a documented rollback plan

Complete inventory of all systems under EJ's oversight. Verify each during nightly startup and Active Projects review.

Core System

Management Dashboards

Operations Tools

Intelligence & Automation

Pending Go/No-Go Decisions (EJ Action Required)