🧪 PRODUCTION TESTING GUIDE

Purpose: Step-by-step manual verification of all MenuListAi intelligence features
Audience: Founder/QA before production launch
Date: January 11, 2026
Estimated Time: 2-3 hours for complete verification

---

📋 PRE-TESTING CHECKLIST

Before starting, ensure:

Requirement	How to Verify	Status
Dev server running	npm run dev at localhost:3000	☐
Firebase connected	Check console for connection logs	☐
Test store exists	At least 1 store with menu data	☐
Menu has 10+ items	Items with images, prices, categories	☐
Analytics data exists	At least 7 days of simulated views/clicks	☐
Feature flags enabled	All flags true in src/config/features.ts	☐

Quick Feature Flag Check

Open src/config/features.ts and verify:

ENABLE_DECISION_BLOCKS: true; // Line 298
SOCIAL_CONTENT_ENABLED: true; // Line 337
DIGITAL_SCREENS_ENABLED: true; // Line 415
MENU_INTELLIGENCE_ENABLED: true; // Line 439

---

FEATURE 1: DECISION BLOCKS

What It Does

Shows 3 smart recommendation blocks at the top of customer menu (Popular, Quick Pick, Best Value).

Where to Test

Customer Menu: https://{subdomain}.menulist.ai or localhost:3000/_client/{subdomain}

Test Cases

TEST 1.1: Decision Blocks Appear

Step	Action	Expected Result
1	Open customer menu URL	Page loads successfully
2	Look at top of menu (below header)	See 3 Decision Block cards
3	Check block titles	"Popular Right Now", "Quick Pick", "Best Value"
4	Verify each block shows item	Name, image, price, reason text

Pass Criteria: All 3 blocks visible with real menu items

TEST 1.2: Runtime Availability Filtering

Step	Action	Expected Result
1	In admin, mark an item as unavailable	Save changes
2	Refresh customer menu	Unavailable item NOT shown in blocks
3	Mark item available again	Item may reappear in blocks

Pass Criteria: Unavailable items never appear in Decision Blocks

TEST 1.3: Click Tracking

Step	Action	Expected Result
1	Open browser dev tools → Network tab	Filter by "analytics"
2	Click on a Decision Block item	See analytics request fire
3	Verify request contains recommendationClicks	Data being tracked

Pass Criteria: Clicks are tracked to analytics

TEST 1.4: Scroll to Item

Step	Action	Expected Result
1	Note the item shown in "Popular" block	Remember item name
2	Click on the block	Page scrolls to that item
3	Observe item highlight	2-second visual highlight

Pass Criteria: Smooth scroll + highlight animation

TEST 1.5: Owner Pin Controls

Step	Action	Expected Result
1	Go to Project Editor → Smart Recommendations	Settings modal opens
2	Pin a specific item to "Popular"	Save settings
3	Refresh customer menu	Pinned item appears in "Popular"

Pass Criteria: Owner pins override AI selection

TEST 1.6: Block Disable

Step	Action	Expected Result
1	In Smart Recommendations, disable "Quick Pick"	Toggle off
2	Refresh customer menu	Only 2 blocks visible
3	Re-enable "Quick Pick"	3 blocks visible again

Pass Criteria: Owner can hide individual blocks

---

Decision Blocks Verification Checklist

Test	Status	Notes
1.1 Blocks Appear	☐
1.2 Availability Filter	☐
1.3 Click Tracking	☐
1.4 Scroll to Item	☐
1.5 Owner Pin	☐
1.6 Block Disable	☐

---

FEATURE 2: CONTINUOUS MENU INTELLIGENCE (CMI)

What It Does

Runs nightly at 2:30 AM UTC, scores every menu item, stores confidence in Firestore.

Where to Test

Admin: Check Firebase Console → Firestore → menuIntelligence collection

Test Cases

TEST 2.1: Intelligence Document Exists

Step	Action	Expected Result
1	Open Firebase Console	Navigate to Firestore
2	Go to menuIntelligence collection	Collection exists
3	Find document: {tId}_{sId}_{projectId}	Document with item scores

Pass Criteria: Document exists with itemScores array

TEST 2.2: Item Scores Structure

Step	Action	Expected Result
1	Open a menuIntelligence document	View data
2	Check itemScores array	Each item has: itemId, confidence, stableDays, trend
3	Verify confidence values	Between 0.0 and 1.0

Pass Criteria: Valid score structure for all items

TEST 2.3: Scheduler Runs (Manual Trigger)

Step	Action	Expected Result
1	Open Firebase Console → Functions	Find decisionBlocksScoring
2	Trigger manually (or wait for 2:30 AM UTC)	Function executes
3	Check logs	"CMI computed" message
4	Check Firestore	lastUpdated timestamp updated

Pass Criteria: Scheduler updates intelligence data

TEST 2.4: Confidence Thresholds Applied

Step	Action	Expected Result
1	Find an item with high views/clicks	Note its confidence
2	Find an item with low views/clicks	Note its confidence
3	Compare values	High-engagement item has higher confidence

Pass Criteria: Confidence correlates with engagement

---

CMI Verification Checklist

Test	Status	Notes
2.1 Document Exists	☐
2.2 Scores Structure	☐
2.3 Scheduler Runs	☐
2.4 Thresholds Applied	☐

---

FEATURE 3: DIGITAL SCREENS

What It Does

Auto-generates slideshow content for in-store TVs. Runs itself with zero owner effort.

Where to Test

Screen URL: localhost:3000/screen/{token} (get token from store settings)

Test Cases

TEST 3.1: Screen Loads

Step	Action	Expected Result
1	Get screen token from store settings	Copy the URL
2	Open screen URL in browser	Screen renders
3	See rotating slides	8-second intervals

Pass Criteria: Screen loads and rotates content

TEST 3.2: Slide Content Hierarchy

Step	Action	Expected Result
1	Watch screen for 1 minute	Observe slide types
2	Check for campaign slides	Items with confidence ≥ 0.7
3	Check for evergreen slides	Meal-based (Breakfast/Lunch/Dinner)
4	Check for brand fallback	Store name/logo slide

Pass Criteria: 4-layer stack (Owner → Campaign → Evergreen → Brand)

TEST 3.3: Zero-Blank Guarantee

Step	Action	Expected Result
1	Create a store with no campaigns	Fresh store
2	Open screen URL	Screen still shows content
3	Verify brand fallback	Store name appears

Pass Criteria: Screen NEVER shows blank

TEST 3.4: Minimum/Maximum Slides

Step	Action	Expected Result
1	Count slides in rotation	Note the count
2	Verify minimum	At least 3 slides
3	Verify maximum	At most 8 slides

Pass Criteria: 3-8 slides always

TEST 3.5: Offline Resilience (Cache)

Step	Action	Expected Result
1	Load screen URL	Wait for first render
2	Open dev tools → Application → localStorage	See menulist-screen-data
3	Disconnect network	Go offline
4	Refresh page	Screen loads from cache

Pass Criteria: Works offline with cached data

TEST 3.6: Real-Time Updates

Step	Action	Expected Result
1	Keep screen open	In one browser tab
2	In admin, trigger campaign regeneration	Save/update data
3	Watch screen	Content updates without refresh

Pass Criteria: Firebase listener updates content

TEST 3.7: Owner Uploads (Max 3)

Step	Action	Expected Result
1	Go to Digital Screen settings	In admin
2	Upload 3 custom images	All upload successfully
3	Try to upload 4th	Blocked with limit message

Pass Criteria: Maximum 3 owner uploads enforced

TEST 3.8: Upload Expiry (14 Days)

Step	Action	Expected Result
1	Upload a custom image	Note the date
2	Check upload metadata	Has expiry timestamp
3	After 14 days	Image auto-removed (or test with mock date)

Pass Criteria: Uploads expire after 14 days

---

Digital Screens Verification Checklist

Test	Status	Notes
3.1 Screen Loads	☐
3.2 Slide Hierarchy	☐
3.3 Zero-Blank	☐
3.4 Min/Max Slides	☐
3.5 Offline Cache	☐
3.6 Real-Time Updates	☐
3.7 Max 3 Uploads	☐
3.8 Upload Expiry	☐

---

FEATURE 4: SOCIAL CONTENT (TODAY TAB)

What It Does

Shows owner ONE action per day. Generates campaigns based on menu intelligence.

Where to Test

Admin: Sidebar → "Today" tab

Test Cases

TEST 4.1: Today Screen Loads

Step	Action	Expected Result
1	Login to admin	Dashboard loads
2	Click "Today" in sidebar	Today screen appears
3	Check for primary card	Campaign or empty state

Pass Criteria: Today screen renders without errors

TEST 4.2: Primary Campaign Card

Step	Action	Expected Result
1	If campaign exists	See PrimaryCard component
2	Check card content	Item image, name, action buttons
3	Check campaign type	One of 9 types (5 active, 4 passive)

Pass Criteria: Primary card shows relevant campaign

TEST 4.3: Empty State (Silence)

Step	Action	Expected Result
1	If no campaign meets confidence	See EmptyState component
2	Check message	"Nothing urgent today" or similar
3	No action required	Silence is valid

Pass Criteria: Graceful empty state (not an error)

TEST 4.4: Campaign Actions

Step	Action	Expected Result
1	On PrimaryCard, click "Share to WhatsApp"	WhatsApp share intent opens
2	Click "Download Poster"	Image downloads
3	Click "Skip"	Campaign dismissed

Pass Criteria: All action buttons work

TEST 4.5: Post-Action State

Step	Action	Expected Result
1	Complete a campaign action	Click "Done"
2	See PostActionState	"Nice work" message
3	Check for outcome message	Non-comparative feedback

Pass Criteria: Post-action feedback shown

TEST 4.6: Operational Section

Step	Action	Expected Result
1	Scroll below primary card	See secondary campaigns
2	Check for passive campaigns	"Today's Special", "Weekend Pick", etc.
3	Verify these are below the fold	Not competing with primary

Pass Criteria: Operational campaigns show as secondary

TEST 4.7: Silence Governor

Step	Action	Expected Result
1	Complete 4+ actions in 7 days	Track in test data
2	Check for intentional silence	Some days show empty state
3	Verify no "always something" trap	System chooses quiet days

Pass Criteria: Active owners get silence days

---

Social Content Verification Checklist

Test	Status	Notes
4.1 Today Loads	☐
4.2 Primary Card	☐
4.3 Empty State	☐
4.4 Campaign Actions	☐
4.5 Post-Action	☐
4.6 Operational Section	☐
4.7 Silence Governor	☐

---

FEATURE 5: PHYSICAL SURFACES

What It Does

Generates downloadable tent cards (PDF) and counter stickers (PNG) for printing.

Where to Test

Admin: Today tab → Physical Surfaces section (when eligible)

Test Cases

TEST 5.1: Tent Card Eligibility

Step	Action	Expected Result
1	Go to Today tab	Check for Tent Card section
2	If confidence ≥ 0.7	Tent Card section visible
3	If confidence < 0.7	Section hidden (not error)

Pass Criteria: Shows only when confident

TEST 5.2: Tent Card Download

Step	Action	Expected Result
1	Click "Download Tent Card"	PDF generates
2	Open downloaded PDF	Valid A6/A5 layout
3	Check content	Item name, QR code, brand footer

Pass Criteria: Valid PDF with correct content

TEST 5.3: Counter Sticker Eligibility

Step	Action	Expected Result
1	Check confidence threshold	Must be ≥ 0.8
2	If eligible	Counter Sticker section visible
3	Higher bar than tent card	Fewer items qualify

Pass Criteria: Only ultra-confident items get stickers

TEST 5.4: Counter Sticker Download

Step	Action	Expected Result
1	Click "Download Sticker"	PNG generates
2	Open downloaded PNG	80mm × 80mm at 300dpi
3	Check content	Copy, QR code, item name

Pass Criteria: Valid PNG with correct size

TEST 5.5: Template Variety

Step	Action	Expected Result
1	Note the template used	System-selected
2	Check template copy	Matches defined templates
3	Verify no Template 5	Banned per spec

Pass Criteria: Only templates 1-4 used

---

Physical Surfaces Verification Checklist

Test	Status	Notes
5.1 Tent Card Eligibility	☐
5.2 Tent Card Download	☐
5.3 Sticker Eligibility	☐
5.4 Sticker Download	☐
5.5 Template Variety	☐

---

FEATURE 6: STAFF PROMPT

What It Does

Shows ONE sentence for staff to repeat when customers ask "What's good?"

Where to Test

Admin: Today tab → Staff Prompt section (rare appearance)

Test Cases

TEST 6.1: Staff Prompt Eligibility (8 Gates)

Step	Action	Expected Result
1	Check primary campaign confidence	Must be ≥ 0.8
2	Check item stability	Must be 10+ days
3	Check prior surface validation	Must have appeared elsewhere
4	If all gates pass	Staff Prompt section visible

Pass Criteria: Only ultra-confident, stable items qualify

TEST 6.2: Sentence Structure

Step	Action	Expected Result
1	Read the prompt text	"Most people take the {itemName}."
2	Verify NO variants	No "Many customers", no "Regulars"
3	Same structure always	Immutable

Pass Criteria: Single sentence structure only

TEST 6.3: Read-Only Display

Step	Action	Expected Result
1	Check for edit buttons	NONE should exist
2	Check for copy buttons	NONE should exist
3	Check for "send to staff"	NONE should exist

Pass Criteria: Completely read-only

TEST 6.4: Inertia Rules

Step	Action	Expected Result
1	Note the item shown today	Record it
2	Check tomorrow	Same item for 3+ days
3	Check weekly appearances	Max 2 days per week

Pass Criteria: Stable, not flickering

TEST 6.5: Rare Appearance

Step	Action	Expected Result
1	Track appearances over 10 days	Record each day
2	Calculate appearance rate	< 30% of days
3	Silence is normal	Not showing = correct

Pass Criteria: Authority through scarcity

---

Staff Prompt Verification Checklist

Test	Status	Notes
6.1 8-Gate Eligibility	☐
6.2 Sentence Structure	☐
6.3 Read-Only	☐
6.4 Inertia Rules	☐
6.5 Rare Appearance	☐

---

CROSS-FEATURE TESTS

Test A: Confidence Hierarchy

Surface	Threshold	Test
Campaign (Passive)	0.3	Items with 0.3+ appear in operational
Campaign (Active)	0.6	Items with 0.6+ become primary
Decision Blocks	0.65	Items with 0.65+ appear in blocks
Digital Screen	0.7	Items with 0.7+ appear on screen
Tent Card	0.7	Items with 0.7+ get tent cards
Counter Sticker	0.8	Items with 0.8+ get stickers
Staff Prompt	0.8	Items with 0.8+ AND 10 days stable

Test: Verify an item with confidence 0.5 appears in campaigns but NOT on screens.

Test B: Data Flow

Step	Action	Expected Result
1	Update menu item in editor	Save changes
2	Check Decision Blocks	Item available status reflected
3	Check Today tab	Campaign updated
4	Check Digital Screen	Slide updated

Pass Criteria: Changes propagate across all surfaces

Test C: Multi-Project Fairness

Step	Action	Expected Result
1	Store with 2+ projects	Both have campaigns
2	Track which project features	Over 7 days
3	Verify rotation	No single project dominates forever

Pass Criteria: Recency decay works

---

FINAL PRODUCTION CHECKLIST

Before Going Live

Category	Item	Status
Code	All tests above pass	☐
Security	withAuth() on all protected routes	☐
Config	Feature flags set correctly	☐
Config	Environment variables in production	☐
Database	Firestore rules deployed	☐
Monitoring	Sentry enabled (optional)	☐
Docs	All spec/impl docs current	☐

Go-Live Sequence

1. Deploy to staging → Run all tests above
2. Enable for 1 beta store → Monitor 24 hours
3. Check Sentry → Zero critical errors
4. Enable for 10 stores → Monitor 7 days
5. Full rollout → Enable for all stores

---

Test Results Summary

Feature	Tests	Passed	Failed
Decision Blocks	6	☐	☐
CMI	4	☐	☐
Digital Screens	8	☐	☐
Social Content	7	☐	☐
Physical Surfaces	5	☐	☐
Staff Prompt	5	☐	☐
Cross-Feature	3	☐	☐
TOTAL	38	☐	☐

---

Sign-Off

╔══════════════════════════════════════════════════════════════════════════════╗
║                                                                                ║
║  TESTER: ____________________                                                  ║
║  DATE:   ____________________                                                  ║
║                                                                                ║
║  ALL 38 TESTS PASSED: [ ] YES  [ ] NO                                          ║
║                                                                                ║
║  PRODUCTION READY: [ ] YES  [ ] NO                                             ║
║                                                                                ║
║  NOTES: _______________________________________________________________        ║
║                                                                                ║
╚══════════════════════════════════════════════════════════════════════════════╝

---

Testing Guide Generated: January 11, 2026
Version: 1.0.0