Workshop 5: Full Stack illustration
Workshop 05Advanced90 minEveryone

Full Stack

Wire everything together into a morning dashboard that runs before you sit down

What you'll learn4 objectives
  • Build a complete morning dashboard that combines overnight scheduled tasks, live MCP queries, and custom skills into a single Word document (.docx) ready for review in 5 minutes
  • Understand multi-agent Dispatch for running parallel skills that produce a combined output
  • Conduct an honest audit of every automation built across Workshops 1-5 with keep/kill/iterate decisions
  • Create a personal 30-day plan that turns workshop exercises into daily working habits
0 of 9
1

Promise Check-In

Intro

Did your scheduled tasks fire this week? One sentence on the best success and one on the biggest failure — both are useful.

2 min

Final check-in. Show of hands.

Best success story -- one sentence. Biggest failure -- one sentence. We want both.

Pattern to listen for: most failures are infrastructure (laptop sleeping, token expired, Claude Desktop not running) rather than prompt issues. If the task ran, it probably produced reasonable output.

Who had scheduled tasks fire successfully this week?
Best success story — one sentence.
Biggest failure — one sentence.
What infrastructure issue caused the most problems?
2

Systems Check

Step-by-Step

Verify all five layers of your stack are live: workspace folder, CLAUDE.md, MCP connectors, skills, and scheduled tasks from Workshop 4.

5 min

Before we build the final piece, make sure everything from Workshops 1-4 is working. Run /sun:setup to verify your stack is intact.

MCP connections

Jira, Confluence, Glean — are they connected and authenticated?

Fix: Settings > Integrations > Reconnect. Re-authenticate via SSO.

Skills installed

Pre-built skills + custom skill from Workshop 4 — do they respond?

Fix: Reinstall from skill marketplace or course skills list.

Scheduled tasks active

2 tasks from Workshop 4 — are they active? Did they produce output?

Fix: Check Claude Desktop was running. Verify workspace path. Re-test on-demand.

CLAUDE.md current

Role, department, tools, terminology — still accurate?

Fix: Edit directly. Add anything learned since Workshop 1.

Things break. Connections expire, laptops sleep through scheduled tasks, files get moved. Diagnosing and recovering is a skill, not a failure. Do a systems check weekly.
3

8:55 AM

Intro

The facilitator opens their laptop at 8:55 AM. Three reports are already waiting in their workspace — generated by scheduled skills while they slept.

3 min

8:55 AM. Three docs are already waiting in the workspace — generated while the facilitator slept. Four-minute review. No tab-hopping. No Slack archaeology.

Today you build YOUR version of this.

The facilitator's morning dashboard was built over 3 weeks of iteration. Your first version today will be simpler. A dashboard that covers 60% of your morning needs is still dramatically better than opening 5 tabs and scrolling through each one.
4

Data Pack Morning Briefing

Exercise

Generate your complete Monday morning briefing from overnight alerts, weekly report, and your department data — the capstone of everything you've built.

5 min

Your data pack has overnight-alerts.xlsx and weekly-summary.md — what your Workshop 4 scheduled tasks would produce. Run the prompt below to see Cowork stitch them into one briefing.

Read overnight-alerts.xlsx and weekly-summary.md. Combine them into a single morning briefing I can review in 5 minutes. Highlight the 3 most urgent items, flag any patterns, and list action items for today.
5

Multi-Agent Dispatch

Step-by-Step

Wire three skills to run in parallel using Dispatch — briefing, verification, and action list produced simultaneously from a single command.

10 min

Three skills, one command, in parallel. Total time = the slowest agent, not the sum.

How it works
  • Agent 1 — Triage overnight alerts
  • Agent 2 — Jira briefing
  • Agent 3 — Yesterday's carry-overs

All three output files land in the workspace. Cowork stitches them into one dashboard.

Run these three tasks in parallel:
1. Triage overnight alerts — save as alerts-{date}.docx
2. Pull Jira briefing — save as briefing-{date}.docx
3. Check yesterday's unfinished items — save as carry-over-{date}.docx

Combine them into morning-dashboard-{date}.docx with an executive summary (3 bullets).
Start simple. Add agents when the bottleneck is sequential processing. A single-source triage is faster with one agent. Three-plus sources benefit from parallel Dispatch.
6

The Honest Audit

Safety

KEEP, KILL, or ITERATE — rate every automation you built across 5 workshops. Calculate your real time savings. Be honest about what's actually working.

10 min

Six automations. Three verdicts. Open audit-checklist.xlsx and score each one honestly — no points for polite answers.

1
Does this actually save time?
When you used it on real data, did it actually take less time than doing it by hand? The honest answer is often '20 minutes, not 2 hours.' That is still worth it at 50 weeks per year.
2
Is the output trustworthy enough to act on?
After your review, would you share this with your manager? Would you use this to make a decision? If you'd rebuild it from scratch anyway, the automation is not saving time.
3
Would you stake your name on this?
If this report goes to a regulator, a client, or a board member — is the quality high enough? If yes, keep. If 'yes, after review,' iterate to reduce review burden. If no, kill.
Reconciliation (Workshop 1)
Cross-system briefing (Workshop 2)
Department triage (Workshop 3)
Custom skill (Workshop 4)
Scheduled tasks (Workshop 4)
Morning dashboard (Workshop 5)
Calculate the ROI for my automation portfolio.
For each task I automated, provide: manual time, Cowork time (generation + review), frequency per week.
Monthly time-savings summary in Word document (.docx).
Distinguish 'time redirected' (same hours, better work) from 'time reclaimed' (actual free time).
Not everything should be automated. The honest answer is sometimes 'this is faster by hand.' Killing an automation that does not work is not failure — it is good judgment.
7

Lightning Demos

Exercise

2 minutes each, strictly enforced. Share your screen and show the room the one workflow you're most proud of from the last 5 weeks.

15 min
2 min per demo — strictly enforced2:00

Six volunteers. Two minutes each. No slides — just screen share and the one workflow you're proud of.

Watch For

  • Steal-worthy ideas from other departments
  • Common patterns across all dashboards (overnight summary, deadline watch, carry-over)
  • Creative solutions you had not considered
2 minutes, strictly enforced
1Open workspace folder, show the morning dashboard file
2Walk through sections — what data does each section pull?
3One thing that surprised you — error caught, pattern found, insight from review
4One sentence: 'The thing that saves me the most time is...'

Most Likely to Actually Get Used on Monday

Not the flashiest — the one that will genuinely be opened every morning

Best Time Saver

The dashboard that replaces the most manual work

Why Didn't I Think of That

The most creative or unexpected approach

The real value of lightning demos is cross-pollination. The chargeback analyst's deadline tracker might inspire your finance close tracker. Steal good ideas and add them to your dashboard in Week 2.
8

The Network Reveal

Step-by-Step

See Nexus Solutions — the merchant hiding across Compliance, Chargebacks, Risk, and Sales from Workshop 3 — and how your MCP connections would surface patterns like this automatically.

5 min

Nexus Solutions showed up in Compliance, Chargebacks, Risk, and Inside Sales — all the same merchant, all the same week. Silos hid that. A cross-system dashboard would have surfaced it in one line.

Nexus Solutions

Compliance, Chargebacks, Risk, Inside Sales

Same merchant flagged across 4 departments — invisible until the cross-department reveal.

The most valuable signals cross department boundaries. A compliance flag, a chargeback spike, a sales stall, and an onboarding delay might all be the same merchant. Silos hide this. Cross-system dashboards reveal it.
9

30-Day Plan

Step-by-Step

Leave with a written 4-week roadmap: which automations to keep, what to build next, and your ROI number for your manager.

7 min

Week 1

Run what you built. Fix what breaks.

  • Run morning dashboard every day
  • Check scheduled tasks every morning — diagnose missing output
  • Run custom skill on real data 3+ times
  • Keep a list of what breaks and what surprises
  • Fix one broken thing

Week 2

Refine and extend.

  • Update CLAUDE.md with Week 1 learnings
  • Improve custom skill — better error handling, tighter constraints
  • Share dashboard or skill with one teammate, get feedback
  • Add one new section to morning dashboard

Week 3

Build something new.

  • Identify and automate a second weekly workflow
  • Start measuring time — /sun:roi-calculator or simple timer
  • Add a second scheduled task cadence
  • Help one colleague set up CLAUDE.md and first skill

Week 4

Honest review.

  • Revisit Keep/Kill/Iterate grid — what changed?
  • Kill anything unused for 2 weeks
  • Calculate actual time savings, share with manager
  • Write one-paragraph summary: before vs after
My automations run daily. My morning dashboard is waiting when I open my laptop. I review it with coffee. I know what needs my attention before my first meeting. The rest of my time goes to the work that actually needs a human brain. And when that work is done — I go outside.
The honest ROI of this course is not '10x productivity.' It is: 'My Monday morning went from 2 hours of grunt work to 20 minutes of review.' That is real. That compounds. And that is enough.

Steal This Prompt

Copy these prompts, customize the file paths to match your workspace, and use them in Cowork. Replace ~ with your actual home directory path if needed.

Accounting Morning Dashboard

Start here
Build a morning dashboard that runs at 7am daily. Pull from Jira: open accounting tickets and month-end tasks. Pull from Confluence: updated accounting procedures and close calenda...
Show full prompt
Build a morning dashboard that runs at 7am daily.
Pull from Jira: open accounting tickets and month-end tasks.
Pull from Confluence: updated accounting procedures and close calendars.
Produce a Word document (.docx) with:
- RED: Items past close deadline
- YELLOW: Reconciliations pending review
- GREEN: Completed items
If any RED items exist past close deadline, draft escalation.
Accounting

Chargeback Morning Dashboard

Build a morning dashboard that runs at 7am daily. Pull from Jira: all cases with status changes overnight. Pull from Confluence: any scheme bulletin updates. Combine with: yesterda...
Show full prompt
Build a morning dashboard that runs at 7am daily.
Pull from Jira: all cases with status changes overnight.
Pull from Confluence: any scheme bulletin updates.
Combine with: yesterday's unresolved items from my workspace.
Produce a Word document (.docx) with:
- RED: Cases < 3 days from deadline
- YELLOW: Cases needing attention today
- GREEN: Cases on track
- New scheme bulletins requiring action
If any RED items exist, draft a separate escalation summary.
Chargebacks

Compliance Morning Dashboard

Build a morning dashboard that runs at 7am daily. Pull from Jira: open CDD cases and EDD reviews due this week. Pull from Confluence: updated regulatory guidance and policy changes...
Show full prompt
Build a morning dashboard that runs at 7am daily.
Pull from Jira: open CDD cases and EDD reviews due this week.
Pull from Confluence: updated regulatory guidance and policy changes.
Search Glean: sanctions screening tool alerts from past 24 hours.
Produce a Word document (.docx) with:
- RED: Cases past SLA or flagged for MLRO escalation
- YELLOW: EDD reviews due within 48 hours
- GREEN: Cases progressing on schedule
- New regulatory updates requiring action
If RED cases > 3, draft MLRO escalation summary.
Compliance / CDD

Country Ops Morning Dashboard

Build a morning dashboard that runs at 7am daily. Pull from Jira: country-level operational issues and merchant escalations. Pull from Confluence: market-specific runbooks and comp...
Show full prompt
Build a morning dashboard that runs at 7am daily.
Pull from Jira: country-level operational issues and merchant escalations.
Pull from Confluence: market-specific runbooks and compliance updates.
Produce a Word document (.docx) with:
- RED: Regulatory deadlines or merchant outages
- YELLOW: Pending local partner approvals
- GREEN: Operations running normally
If any market has a regulatory deadline within 7 days, flag separately.
Country Operations

Support Morning Dashboard

Build a morning dashboard that runs at 7am daily. Pull from Jira: overnight escalations and SLA breaches. Pull from Confluence: updated runbooks and known issues. Search Glean: rel...
Show full prompt
Build a morning dashboard that runs at 7am daily.
Pull from Jira: overnight escalations and SLA breaches.
Pull from Confluence: updated runbooks and known issues.
Search Glean: relevant announcements from past 24 hours.
Produce a Word document (.docx) with:
- RED: SLA breaches requiring immediate action
- YELLOW: Escalations pending response
- GREEN: Resolved overnight items
- Trending issues (same category appearing 3+ times)
If SLA breaches > 5, draft escalation to team lead.
Customer Relations

Data Analytics Morning Dashboard

Build a morning dashboard that runs at 7am daily. Pull from Jira: open data requests and pipeline incidents. Pull from Confluence: Snowflake migration status and dashboard change l...
Show full prompt
Build a morning dashboard that runs at 7am daily.
Pull from Jira: open data requests and pipeline incidents.
Pull from Confluence: Snowflake migration status and dashboard change log.
Produce a Word document (.docx) with:
- RED: Pipeline failures or data quality alerts from overnight
- YELLOW: Ad-hoc requests pending > 24 hours
- GREEN: Pipelines healthy, dashboards current
If any pipeline failure occurred overnight, include error summary.
Data / Analytics

Finance Morning Dashboard

Build a morning dashboard that runs at 7am daily. Pull from Jira: settlement exceptions and financial correction requests. Pull from Confluence: month-end close checklist progress....
Show full prompt
Build a morning dashboard that runs at 7am daily.
Pull from Jira: settlement exceptions and financial correction requests.
Pull from Confluence: month-end close checklist progress.
Produce a Word document (.docx) with:
- RED: Unreconciled settlements past 48 hours
- YELLOW: Financial corrections awaiting four-eye approval
- GREEN: Settlements matched and processed
If unreconciled amount > threshold, draft escalation to finance lead.
Finance

Fraud Morning Dashboard

Build a morning dashboard that runs at 7am daily. Pull from Jira: overnight fraud alerts and monitoring case queue. Pull from Confluence: updated fraud pattern bulletins. Produce a...
Show full prompt
Build a morning dashboard that runs at 7am daily.
Pull from Jira: overnight fraud alerts and monitoring case queue.
Pull from Confluence: updated fraud pattern bulletins.
Produce a Word document (.docx) with:
- RED: High-severity alerts (score > 90) requiring immediate review
- YELLOW: Merchants approaching VAMP/MATCH thresholds
- GREEN: Cleared false positives from overnight
- BIN attack or velocity spike patterns detected
If any merchant exceeds 0.9% chargeback ratio, flag for scheme review.
Fraud

Inside Sales Morning Dashboard

Build a morning dashboard that runs at 7am daily. Pull from Jira: CDD status updates on pending merchant applications. Search Glean: overnight inbound leads and HubSpot deal card c...
Show full prompt
Build a morning dashboard that runs at 7am daily.
Pull from Jira: CDD status updates on pending merchant applications.
Search Glean: overnight inbound leads and HubSpot deal card changes.
Produce a Word document (.docx) with:
- RED: Deals at risk of stalling (no activity > 5 days)
- YELLOW: CDD pending items blocking onboarding handoff
- GREEN: Deals progressing through pipeline
- Today's call list with pre-call context
If any deal is past SLA for CDD response, flag for follow-up.
Inside Sales

IT Helpdesk Morning Dashboard

Build a morning dashboard that runs at 7am daily. Pull from Jira: open IT tickets, overnight incidents, and SLA status. Pull from Confluence: known issues and maintenance windows. ...
Show full prompt
Build a morning dashboard that runs at 7am daily.
Pull from Jira: open IT tickets, overnight incidents, and SLA status.
Pull from Confluence: known issues and maintenance windows.
Produce a Word document (.docx) with:
- RED: P1/P2 incidents still open from overnight
- YELLOW: Tickets approaching SLA breach
- GREEN: Resolved overnight items
If any P1 incident is open, draft all-hands notification.
IT / Helpdesk

Legal Morning Dashboard

Build a morning dashboard that runs at 7am daily. Pull from Jira: open legal review requests and contract approvals. Pull from Confluence: regulatory update tracker and policy chan...
Show full prompt
Build a morning dashboard that runs at 7am daily.
Pull from Jira: open legal review requests and contract approvals.
Pull from Confluence: regulatory update tracker and policy changes.
Produce a Word document (.docx) with:
- RED: Contract reviews past deadline
- YELLOW: Regulatory changes requiring legal assessment
- GREEN: Reviews completed and approved
If any regulatory change impacts merchant agreements, flag for review.
Legal

Logistics Morning Dashboard

Build a morning dashboard that runs at 7am daily. Pull from Jira: terminal delivery and swap requests pending. Search Glean: Zuper dispatch status and field engineer availability. ...
Show full prompt
Build a morning dashboard that runs at 7am daily.
Pull from Jira: terminal delivery and swap requests pending.
Search Glean: Zuper dispatch status and field engineer availability.
Produce a Word document (.docx) with:
- RED: Failed deliveries or missed SLA windows
- YELLOW: Deliveries scheduled today needing confirmation
- GREEN: Completed installations and swaps
If any delivery failed twice, escalate to logistics manager.
Logistics / Supply Chain

Marketing Morning Dashboard

Build a morning dashboard that runs at 7am daily. Pull from Jira: content calendar items due this week. Pull from Confluence: campaign briefs and brand guidelines updates. Search G...
Show full prompt
Build a morning dashboard that runs at 7am daily.
Pull from Jira: content calendar items due this week.
Pull from Confluence: campaign briefs and brand guidelines updates.
Search Glean: competitor announcements from past 24 hours.
Produce a Word document (.docx) with:
- RED: Content past publication deadline
- YELLOW: Campaigns launching within 48 hours needing approval
- GREEN: Scheduled content on track
If any competitor launched a relevant feature, draft competitive brief.
Marketing

Onboarding Morning Dashboard

Build a morning dashboard that runs at 7am daily. Pull from Jira: merchant applications in pipeline and CDD clearance status. Pull from Confluence: onboarding process checklists. P...
Show full prompt
Build a morning dashboard that runs at 7am daily.
Pull from Jira: merchant applications in pipeline and CDD clearance status.
Pull from Confluence: onboarding process checklists.
Produce a Word document (.docx) with:
- RED: Applications stalled > 5 business days
- YELLOW: CDD pending items blocking activation
- GREEN: Applications progressing normally
- Conversion funnel: application to first settlement
If stalled applications > 10, draft bottleneck analysis.
Onboarding

Operations Morning Dashboard

Build a morning dashboard that runs at 7am daily. Pull from Jira: operational incidents and process exceptions. Pull from Confluence: updated SOPs and operational playbooks. Produc...
Show full prompt
Build a morning dashboard that runs at 7am daily.
Pull from Jira: operational incidents and process exceptions.
Pull from Confluence: updated SOPs and operational playbooks.
Produce a Word document (.docx) with:
- RED: Active incidents or system outages
- YELLOW: Process exceptions needing manual intervention
- GREEN: Operations running within normal parameters
If any system is degraded, include impact assessment.
Operations

Partnerships Morning Dashboard

Build a morning dashboard that runs at 7am daily. Pull from Jira: partner integration requests and referral pipeline. Pull from Confluence: partner programme documentation updates....
Show full prompt
Build a morning dashboard that runs at 7am daily.
Pull from Jira: partner integration requests and referral pipeline.
Pull from Confluence: partner programme documentation updates.
Produce a Word document (.docx) with:
- RED: Partner SLA breaches or integration failures
- YELLOW: Referral deals pending partner confirmation
- GREEN: Active integrations running normally
If any partner integration has been down > 24 hours, draft escalation.
Partnerships

People / HR Morning Dashboard

Build a morning dashboard that runs at 7am daily. Pull from Jira: open HR tickets and employee requests. Pull from Confluence: policy updates and onboarding checklists. Produce a W...
Show full prompt
Build a morning dashboard that runs at 7am daily.
Pull from Jira: open HR tickets and employee requests.
Pull from Confluence: policy updates and onboarding checklists.
Produce a Word document (.docx) with:
- RED: Employee issues past response SLA
- YELLOW: New starter onboarding items due this week
- GREEN: Resolved requests and completed onboardings
If any compliance training is overdue for > 5 employees, flag.
People / HR

Product Activation Morning Dashboard

Build a morning dashboard that runs at 7am daily. Pull from Jira: merchants pending feature activation and adoption blockers. Pull from Confluence: product release notes and activa...
Show full prompt
Build a morning dashboard that runs at 7am daily.
Pull from Jira: merchants pending feature activation and adoption blockers.
Pull from Confluence: product release notes and activation playbooks.
Produce a Word document (.docx) with:
- RED: Merchants past activation SLA (onboarded but not transacting)
- YELLOW: Feature rollouts pending merchant confirmation
- GREEN: Successfully activated merchants this week
If activation rate drops below target, include root cause analysis.
Product Activation

Product Management Morning Dashboard

Build a morning dashboard that runs at 7am daily. Pull from Jira: sprint board status, blocked tickets, and new bugs. Pull from Confluence: PRD review comments and stakeholder feed...
Show full prompt
Build a morning dashboard that runs at 7am daily.
Pull from Jira: sprint board status, blocked tickets, and new bugs.
Pull from Confluence: PRD review comments and stakeholder feedback.
Produce a Word document (.docx) with:
- RED: Blocked tickets or sprint-at-risk items
- YELLOW: PRDs awaiting stakeholder sign-off
- GREEN: Sprint items on track
- Upcoming: items due for WBR/QBR preparation
If sprint velocity is trending below target, flag.
Product Management

Product Ops Morning Dashboard

Build a morning dashboard that runs at 7am daily. Pull from Jira: product incidents, feature flag changes, and rollback requests. Pull from Confluence: release calendar and operati...
Show full prompt
Build a morning dashboard that runs at 7am daily.
Pull from Jira: product incidents, feature flag changes, and rollback requests.
Pull from Confluence: release calendar and operational readiness checklists.
Produce a Word document (.docx) with:
- RED: Active product incidents or failed deployments
- YELLOW: Releases scheduled for today needing go/no-go
- GREEN: Systems stable, no pending rollbacks
If any deployment failed overnight, include rollback status.
Product Operations

Retention Morning Dashboard

Build a morning dashboard that runs at 7am daily. Pull from Jira: cancellation requests and win-back cases. Search Glean: merchant churn signals from support tickets. Produce a Wor...
Show full prompt
Build a morning dashboard that runs at 7am daily.
Pull from Jira: cancellation requests and win-back cases.
Search Glean: merchant churn signals from support tickets.
Produce a Word document (.docx) with:
- RED: High-value merchants requesting cancellation
- YELLOW: Merchants with declining transaction volumes
- GREEN: Successful win-backs and renewed contracts
If any merchant with MRR > threshold requests cancellation, flag immediately.
Retention

Risk Morning Dashboard

Build a morning dashboard that runs at 7am daily. Pull from Jira: open risk assessments and merchant risk reviews. Pull from Confluence: risk appetite framework and threshold updat...
Show full prompt
Build a morning dashboard that runs at 7am daily.
Pull from Jira: open risk assessments and merchant risk reviews.
Pull from Confluence: risk appetite framework and threshold updates.
Produce a Word document (.docx) with:
- RED: Merchants exceeding risk thresholds or pending suspension
- YELLOW: Risk reviews due within 48 hours
- GREEN: Merchants within acceptable risk parameters
If any merchant triggers multiple risk indicators, draft review summary.
Risk

Scheme Ops Morning Dashboard

Build a morning dashboard that runs at 7am daily. Pull from Jira: scheme compliance items and bulletin action items. Pull from Confluence: Visa/Mastercard/AMEX bulletin summaries. ...
Show full prompt
Build a morning dashboard that runs at 7am daily.
Pull from Jira: scheme compliance items and bulletin action items.
Pull from Confluence: Visa/Mastercard/AMEX bulletin summaries.
Produce a Word document (.docx) with:
- RED: Scheme deadlines within 7 days
- YELLOW: New scheme bulletins requiring action
- GREEN: Compliance items on track
If any scheme mandate has an approaching deadline, draft impact assessment.
Scheme Operations

Treasury Morning Dashboard

Build a morning dashboard that runs at 7am daily. Pull from Jira: settlement funding items and liquidity alerts. Pull from Confluence: treasury policy updates and banking partner n...
Show full prompt
Build a morning dashboard that runs at 7am daily.
Pull from Jira: settlement funding items and liquidity alerts.
Pull from Confluence: treasury policy updates and banking partner notices.
Produce a Word document (.docx) with:
- RED: Funding shortfalls or failed settlement sweeps
- YELLOW: Accounts approaching balance thresholds
- GREEN: All funding positions within normal range
If any settlement funding gap exceeds threshold, draft escalation to CFO.
Treasury

Key Takeaways

  1. Morning dashboard beats tab-hopping — A dashboard covering 60% of your morning needs is dramatically better than opening 5 tabs and scrolling. Start simple; add sources only when the current dashboard is stable.

  2. Scheduled tasks + MCP + custom skills compound — Individually each one saves minutes. Wired together into a morning dashboard, they save an hour before 9am. The value is in the chain, not the links.

  3. Multi-agent Dispatch is for 3+ sources, not bragging rights — A single-source triage is faster with one agent. Parallel agents pay off when the bottleneck is sequential querying across distinct systems.

  4. Honest audit > enthusiastic adoption — Kill anything you have not used in 2 weeks. Iterate anything that works but still needs hand-holding. Keep only what saves real time on real work. Your future self will thank you.

  5. ROI is time redirected, not time saved — Be honest: the hour you freed goes somewhere. Either to higher-value work (name it) or to a shorter day (name it). Unnamed time savings evaporate within a week.

  6. The course is over; the habit isn't — Workshop 1 was 90 minutes. The next 30 days are 30 hours of practice. Pick one automation to run tomorrow morning, with coffee. That is where the outdoor time comes from.

Materials

Select your department using the picker in the top right to see the exercise files for this workshop. Click any file to preview it in place.

Safety Note

A morning dashboard runs while you are still asleep and greets you with a polished document. That polish is the danger — a trustworthy-looking summary based on stale, partial, or hallucinated data will lead to worse decisions than having no summary at all.

Rules that apply from this workshop forward:

  • Never treat a scheduled dashboard as ground truth. It is a draft, even when it looks final. Spot-check at least one number or claim against the source system every morning for the first two weeks.
  • Never forward a morning dashboard — or any extract from it — to a manager, merchant, or external party without reading it end-to-end first
  • Always check the timestamp on every source file the dashboard consumes. A dashboard built from yesterday's alerts and last week's Jira data is worse than no dashboard.
  • Always include a "If this is empty, the task failed" sentinel in every scheduled task so silent failure produces a visible error document, not a plausible-looking empty report
  • Always review your automation portfolio honestly at the end of each month. Kill anything unused. Iterate anything that still needs a human tweak. Keep only what survives real work on real data.