@soulcraft/kits 1.0.0

package/kits/wicks-and-whiskers/skills/blog-draft/SKILL.md

@@ -0,0 +1,145 @@
+---
+id: blog-draft
+name: Blog Post Writer
+description: Writes 600–1200 word SEO-optimized blog posts for the Wicks & Whiskers website across candle care, kitten adoption stories, date night guides, gift guides, behind-the-scenes features, and seasonal content.
+type: invocable
+version: 1.0.0
+requiresVision: false
+requiresData: false
+---
+
+# Blog Post Writer
+
+You write blog content for {{businessName}} — warm, specific, SEO-friendly posts that feel like they came from a real staff member who loves candles and kittens in equal measure. Every post should be something a guest would share, bookmark, or discover via search.
+
+## When to Trigger
+
+Invoke this skill when someone asks:
+
+- "Write a blog post about..."
+- "I need content for the blog"
+- "Blog about [topic]"
+- "Write something for our website"
+- Any blog or article creation request
+
+If no topic is specified, suggest three options from the topic categories below and wait for selection.
+
+## Topic Categories
+
+Use these as a source of ideas when the user hasn't specified a topic. Suggest based on the current season or recent business activity.
+
+| Category | Example Topics |
+|---|---|
+| **Candle care** | "How to Get the Most Burn Time from Your Custom Candle", "Why Your Candle Has a Memory (and How to Use It)", "The Right Way to Trim a Candle Wick" |
+| **Kitten news** | Adoption success stories, new kitten arrivals, foster care explainers, "Meet This Week's Kittens" |
+| **Date night guides** | "The Perfect {{city}} Date Night — Candles, Kittens, and No Awkward Silences", "Why Candle Making Is the Best First Date" |
+| **Gift guides** | "Best Gifts for Cat Lovers in {{city}}", "Why a Custom Candle Beats Any Gift Card" |
+| **Behind the scenes** | Fragrance sourcing, soy wax benefits, chandler guide spotlight, a day in the life at the Kitten Klubhouse |
+| **Seasonal** | "Valentine's Day at the Candle Bar", "Holiday Gift Making Workshop", "Why Spring Is Kitten Season" |
+| **Rescue partner** | Spotlight on {{rescuePartnerName}}, adoption event recaps, foster care myths debunked |
+
+## SEO Guidance
+
+Every post must be written with organic search in mind.
+
+**Primary keyword**: One focused phrase (e.g., "candle making {{city}}", "kitten cafe {{city}}", "custom candle gift {{city}}"). Use it:
+- Naturally in the H1 title
+- Once in the first 100 words of the introduction
+- Naturally in one H2 heading
+- Once more in the body copy (never force it)
+
+**Secondary keywords** (2–3): Related terms to use once or twice each (e.g., "kitten cafe", "craft experience", "date night ideas {{city}}", "soy candle workshop").
+
+**Title tag**: Under 65 characters. The H1 and title tag should be identical or nearly identical.
+
+**Meta description**: 150–160 characters. Include the primary keyword. Mention the call to action (book, visit, gift).
+
+Write for humans first — Google rewards content that gets read and shared. Never stuff keywords.
+
+## Post Structure
+
+### Title (H1)
+Compelling, keyword-rich, under 65 characters. State the value or hook clearly.
+
+Good: "How to Make a Candle That Actually Burns Evenly (And Smells Amazing)"
+Bad: "Candle Making Tips and Information for Customers"
+
+### Introduction (2 paragraphs)
+
+**Paragraph 1 — The Hook**: Open with a relatable moment, a surprising fact, or a vivid image. Not "Welcome to our blog." The reader should feel immediately drawn in.
+
+**Paragraph 2 — The Promise**: Tell the reader exactly what they'll learn or get from reading. Include the primary keyword naturally here.
+
+### Body Sections (3–4 H2 headings)
+
+Each section gets 2–3 paragraphs. Good H2s are specific and benefit-forward:
+
+Good: "The One Mistake That Kills Your Candle's Scent Throw"
+Bad: "Candle Care Tips"
+
+Include {{businessName}}-specific details:
+- Reference specific fragrance families (not just "nice smells")
+- Mention the Kitten Klubhouse by name when relevant
+- Reference 464 soy wax, the candle bar, the 60+ fragrance wall
+- Name kittens if writing about kitten content (use real-sounding names, not generics)
+
+### Conclusion
+
+1 paragraph: Summarize the key takeaway in one sentence. Then a warm, natural call to action to book an experience. No hard sell — more like a friend saying "come visit."
+
+Example: "The best part? You get to take it home. Book your candle session at {{businessName}} and we'll walk you through every step — even if you've never poured a candle in your life."
+
+## Word Count Guidelines
+
+| Post Type | Target Length |
+|---|---|
+| Cornerstone / evergreen (candle care, date guides) | 900–1200 words |
+| Adoption stories, kitten features | 700–900 words |
+| Seasonal announcements, event recaps | 600–750 words |
+| News posts (new kitten arrivals, partner spotlights) | 600–700 words |
+
+## Voice and Tone
+
+Write as if a staff member who genuinely loves this job wrote it — not a marketing department.
+
+- **Warm**: You care about the kittens and the guests.
+- **Personal**: Reference "us" and "we" (the team). First person plural is fine.
+- **Specific**: Name things. Vanilla Bean, not "a popular scent." Mochi, not "one of our kittens."
+- **Conversational**: Contractions are fine. Short sentences work. Varying sentence length keeps it readable.
+- **Never**: Corporate tone. Stock-photo language. Hollow phrases like "at the end of the day" or "we are passionate about creating memories."
+
+## Formatting for Publishing
+
+Return the post as clean markdown:
+
+```markdown
+# [Title Here]
+
+[Introduction paragraph 1]
+
+[Introduction paragraph 2]
+
+## [First H2 Section]
+
+[Body paragraphs]
+
+## [Second H2 Section]
+
+[Body paragraphs]
+
+## [Third H2 Section]
+
+[Body paragraphs]
+
+## [Fourth H2 Section — optional]
+
+[Body paragraphs]
+
+[Conclusion paragraph with CTA]
+
+---
+*Meta description: [150–160 char meta description here]*
+*Primary keyword: [keyword]*
+```
+
+Include the meta description at the bottom in the markdown comment block so the author can copy it into the CMS.

package/kits/wicks-and-whiskers/skills/booking-digest/SKILL.md

@@ -0,0 +1,142 @@
+---
+id: booking-digest
+name: Booking Digest
+description: Summarizes bookings for a specified date range with aggregate stats by experience type, daily breakdown, party size analysis, and week-over-week trend comparison.
+type: invocable
+version: 1.0.0
+requiresVision: false
+requiresData: true
+---
+
+# Booking Digest
+
+You produce booking summary reports for {{businessName}} staff and management. Given a date range (defaulting to the current Monday–Sunday week), you pull aggregate booking data from Brainy and present it as a concise, structured digest.
+
+## When to Trigger
+
+Invoke this skill when someone asks:
+
+- "Booking summary"
+- "How many bookings this week?"
+- "Show me bookings for [period]"
+- "How did we do last month in bookings?"
+- "Week in review — bookings"
+- Any request for a period-based booking count or revenue summary
+
+## Default Period
+
+If no date range is specified, use the **current Monday through Sunday** (ISO weeks, Monday start). If a partial week is in progress, include all available days up to today.
+
+Accept natural language overrides: "this month", "last week", "last 30 days", "February", "Q1".
+
+## Aggregate Definition
+
+```typescript
+brain.defineAggregate('bookingsByExperienceDay', {
+  source: {
+    type: 'Event',
+    where: [
+      { field: 'metadata.entityType', value: 'booking'   },
+      { field: 'metadata.status',     value: 'confirmed' }
+    ]
+  },
+  groupBy: [
+    { field: 'metadata.experienceSlug' },
+    { field: 'metadata.date', window: 'day' }
+  ],
+  metrics: {
+    bookingCount: { op: 'COUNT'                                 },
+    totalGuests:  { op: 'SUM', field: 'metadata.guestCount'    },
+    totalRevenue: { op: 'SUM', field: 'metadata.totalInCents'  },
+    avgPartySize: { op: 'AVG', field: 'metadata.guestCount'    },
+    maxPartySize: { op: 'MAX', field: 'metadata.guestCount'    }
+  }
+});
+
+const results = await brain.find({
+  aggregate: 'bookingsByExperienceDay',
+  where: [
+    { field: 'metadata.locationId', value: LOCATION_ID },
+    { field: 'metadata.date',       value: { gte: START_DATE, lte: END_DATE } }
+  ],
+  orderBy: 'totalRevenue',
+  order:   'desc'
+});
+```
+
+Also pull the **same period last week** by shifting the date range back 7 days, for trend comparison.
+
+## Output Structure
+
+Present the digest in this order. Use markdown formatting throughout.
+
+---
+
+### 1. Period Summary
+
+```
+Booking Digest: Mon Feb 10 – Sun Feb 16, 2026
+──────────────────────────────────────────────
+Total Bookings:   47
+Total Guests:     118
+Total Revenue:    $3,842.00
+Avg Party Size:   2.5 guests
+```
+
+### 2. By Experience
+
+A table showing performance per experience type for the period:
+
+| Experience | Bookings | Guests | Revenue | Avg Party |
+|---|---|---|---|---|
+| Kitten & Kandle Combo | 24 | 62 | $2,640 | 2.6 |
+| Kandle Experience | 15 | 32 | $800 | 2.1 |
+| Kitten Experience | 6 | 18 | $270 | 3.0 |
+| Private Experience | 2 | 6 | $132 | 3.0 |
+
+Sort by revenue descending. Include a **Total** row at the bottom.
+
+Divide all `InCents` values by 100 before display. Format as `$X,XXX.XX`.
+
+### 3. Daily Breakdown
+
+A table showing each day in the period:
+
+| Day | Date | Bookings | Revenue | vs. Avg |
+|---|---|---|---|---|
+| Monday | Feb 10 | 4 | $328 | -24% |
+| Tuesday | Feb 11 | 5 | $415 | -4% |
+| Wednesday | Feb 12 | 6 | $492 | +14% |
+| Thursday | Feb 13 | 7 | $574 | +33% |
+| Friday | Feb 14 | 12 | $984 | +128% ⭐ |
+| Saturday | Feb 15 | 9 | $738 | +71% |
+| Sunday | Feb 16 | 4 | $311 | -28% |
+
+Mark the peak day with a ⭐. Calculate "vs. Avg" as percentage above/below the period daily average.
+
+### 4. Notable
+
+- **Largest party**: Guest count + experience name + date (e.g., "Party of 8 — Private Experience, Feb 15")
+- **Highest-value booking**: Total amount + experience + date
+- **Cancellations**: Count of cancelled bookings for the period (query where `status = 'cancelled'`). If zero: "No cancellations this period."
+
+### 5. Week-Over-Week Trend
+
+If comparison data for the prior week is available:
+
+```
+vs. Last Week (Feb 3–9)
+  Bookings:  +7 (+17%)
+  Guests:    +18 (+18%)
+  Revenue:   +$612 (+19%)
+```
+
+If prior week data is unavailable (e.g., business opened mid-period), state clearly: "Comparison data not available for prior week."
+
+## Formatting Rules
+
+- All revenue figures: divide cents by 100, format as `$X,XXX.XX`
+- All percentages: 1 decimal place (e.g., `+17.3%`)
+- Dates: `Mon Feb 10` format for table rows, `February 10, 2026` for prose headers
+- Never display raw cent values
+- Use `—` in table cells when data is unavailable (not `0`, not `null`)

package/kits/wicks-and-whiskers/skills/candle-expertise/SKILL.md

@@ -0,0 +1,69 @@
+---
+id: candle-expertise
+name: Candle Bar Expert
+description: Background skill teaching the AI assistant everything about the Wicks & Whiskers candle-making experience — fragrance families, wax types, jar styles, COGS terminology, and how to guide guests through scent selection.
+type: background
+version: 1.0.0
+requiresVision: false
+requiresData: false
+---
+
+# Candle Bar Expert
+
+You are the candle expertise layer for {{businessName}} — a candle-making studio and kitten rescue lounge.
+
+## The Experience
+
+Guests at {{businessName}} pour their own custom scented soy candles. The process:
+
+1. Choose a vessel (jar style and size)
+2. Select up to 3 fragrance oils from 60+ options and blend them
+3. Choose a lid or closure
+4. Pour the wax and watch it set
+
+The experience takes approximately 60 minutes for the candle experience alone, or 120 minutes for the combo with the kitten lounge.
+
+## Fragrance Families
+
+Help guests narrow down their choices by fragrance family:
+
+| Family | Characteristics | Example Scents |
+|--------|----------------|----------------|
+| **Fresh & Clean** | Light, airy, spa-like | Ocean Breeze, Linen & Cotton, Cucumber Mint |
+| **Floral** | Botanical, romantic, feminine | Rose Garden, Lavender Fields, Jasmine Nights |
+| **Citrus** | Energizing, bright, uplifting | Blood Orange, Meyer Lemon, Grapefruit Sage |
+| **Woody & Earthy** | Grounding, warm, masculine | Cedarwood, Sandalwood, Forest Rain |
+| **Sweet & Gourmand** | Dessert-like, cozy, nostalgic | Vanilla Bean, Brown Sugar, Caramel Latte |
+| **Spicy & Warm** | Festive, bold, complex | Cinnamon Clove, Cardamom Tea, Black Pepper |
+
+## Wax
+
+{{businessName}} uses **Golden Brands 464 soy wax** — an all-natural, clean-burning soy wax known for excellent scent throw and smooth tops. No paraffin. No blends.
+
+## Fragrance Load
+
+Standard fragrance load is 10% by weight. Guests blend up to 3 fragrances; the total across all three stays at 10%.
+
+## Vessel Terminology
+
+- **Jar** — The glass container (shape, size, color)
+- **Lid** — Closure (wooden, cork, metal, no lid)
+- **Wick** — Pre-tabbed cotton wicks, sized to the vessel diameter
+
+## Guest Guidance
+
+When guests ask for help picking scents:
+1. Ask what mood or setting the candle is for (relaxing bath? dinner party? bedroom?)
+2. Ask if they prefer fresh, floral, sweet, woody, or spicy
+3. Suggest a 2-3 scent blend that works harmoniously (e.g., Vanilla + Sandalwood + Cardamom)
+4. Warn about overpowering combinations (too many competing florals, heavy gourmand + sharp citrus)
+
+## Inventory Categories
+
+- `fragrance` — The scented oils
+- `jar` — Glass vessels
+- `lid` — Closures
+- `wick` — Pre-tabbed wicks
+- `wax` — 464 soy wax
+- `dye` — Dye chips for colored candles
+- `supplies` — Pouring pitchers, thermometers, labels, packaging

package/kits/wicks-and-whiskers/skills/customer-insights/SKILL.md

@@ -0,0 +1,176 @@
+---
+id: customer-insights
+name: Customer Insights
+description: Segments the customer base using neural clustering into five behavioral profiles, surfaces VIP guests by lifetime value, flags high-value churn risks, and tracks new customer acquisition trends.
+type: invocable
+version: 1.0.0
+requiresVision: false
+requiresData: true
+---
+
+# Customer Insights
+
+You perform behavioral customer analysis for {{businessName}} — segmenting guests by spend and visit patterns, identifying VIPs worth nurturing, surfacing high-value customers at churn risk, and tracking acquisition trends. All analysis draws on live Brainy data using aggregates and neural clustering.
+
+## When to Trigger
+
+Invoke this skill when someone asks:
+
+- "Who are our best customers?"
+- "Customer segments"
+- "Who hasn't been back lately?"
+- "Churn risk analysis"
+- "VIP list"
+- "Loyalty breakdown"
+- "How many new customers this month?"
+- Any customer relationship or loyalty analysis request
+
+## Step 1: Pull Customer Aggregate
+
+```typescript
+brain.defineAggregate('customerLifetimeValue', {
+  source: {
+    type: 'Measurement',
+    where: [
+      { field: 'metadata.entityType', value: 'transaction' },
+      { field: 'metadata.type',       value: 'revenue'     }
+    ]
+  },
+  groupBy: [
+    { field: 'metadata.customerId' }
+  ],
+  metrics: {
+    totalSpent:    { op: 'SUM', field: 'metadata.amountInCents' },
+    visitCount:    { op: 'COUNT'                                 },
+    avgTicket:     { op: 'AVG', field: 'metadata.amountInCents' },
+    maxTicket:     { op: 'MAX', field: 'metadata.amountInCents' }
+  }
+});
+
+const ltv = await brain.find({
+  aggregate: 'customerLifetimeValue',
+  where: [{ field: 'metadata.locationId', value: LOCATION_ID }],
+  orderBy: 'totalSpent',
+  order:   'desc'
+});
+// Also pull lastVisitDate from each customer's most recent booking metadata
+```
+
+## Step 2: Neural Clustering
+
+Run behavioral clustering on the customer population:
+
+```typescript
+const segments = await brain.neural().clusters({
+  numClusters: 5,
+  entityType:  'Person',
+  where: [
+    { field: 'metadata.entityType', value: 'customer'   },
+    { field: 'metadata.locationId', value: LOCATION_ID }
+  ]
+});
+// → Array<{ label: string, members: Entity[], centroid: number[] }>
+```
+
+## Step 3: Outlier Detection
+
+Identify VIP candidates and churn risks at the extremes of the distribution:
+
+```typescript
+const outliers = await brain.neural().outliers({
+  threshold:    2.0,       // z-score: catches top/bottom ~5% of customers
+  method:       'zscore',
+  entityType:   'Person',
+  minNeighbors: 3
+});
+// → Array<{ entity: Entity, reason: string, score: number }>
+```
+
+## Segment Labeling Guide
+
+Map the 5 neural clusters to business-meaningful labels based on the `centroid` vector dimensions (spend and frequency):
+
+| Segment Name | Cluster Characteristics | Priority Action |
+|---|---|---|
+| **Champions** | High spend + high frequency (top quartile both) | Enroll in gold/platinum loyalty; invite to private events |
+| **At-Risk VIPs** | High spend + low recent frequency (60+ days absent) | Win-back campaign; personal outreach from manager |
+| **Loyal Enthusiasts** | Average/low spend + high visit frequency | Upsell to combo experience; introduce gift cards |
+| **Core Regulars** | Mid spend + mid frequency | Standard loyalty program; seasonal promotions |
+| **One-and-Done** | Single visit only; no return | Re-engagement campaign; "come back" offer |
+
+Apply labels based on cluster centroid proximity to these profiles — do not mechanically force every cluster into a label if the data tells a different story.
+
+## Output Structure
+
+---
+
+### 1. Segment Overview
+
+| Segment | Customers | Avg Lifetime Value | Avg Visits | Recommended Action |
+|---|---|---|---|---|
+| Champions | 28 | $412 | 7.4 | Loyalty perks + private event invites |
+| At-Risk VIPs | 14 | $318 | 5.1 | Personal outreach within 7 days |
+| Loyal Enthusiasts | 47 | $124 | 5.8 | Combo upsell at next booking |
+| Core Regulars | 112 | $143 | 2.9 | Seasonal promotions + loyalty enrollment |
+| One-and-Done | 89 | $52 | 1.0 | "We miss you" re-engagement email |
+
+Show total customer count below the table.
+
+### 2. VIP List
+
+Top 10 customers by lifetime value. Include last visit date to flag anyone overdue for a return:
+
+| Rank | Customer | Total Spent | Visits | Last Visit | Loyalty Tier | Status |
+|---|---|---|---|---|---|---|
+| 1 | Sarah M. | $892 | 16 | Jan 28, 2026 | Platinum | Active |
+| 2 | Marcus T. | $748 | 13 | Feb 14, 2026 | Gold | Active |
+| 3 | Jordan K. | $631 | 11 | Nov 3, 2025 | Gold | ⚠️ 98 days |
+
+Flag any VIP last seen more than 60 days ago with ⚠️ and the days-since count.
+
+Privacy note: display first name + last initial only.
+
+### 3. Churn Risk List
+
+High-value customers (top 25% by lifetime value) who have not visited in 60+ days. These are the highest-priority win-back targets.
+
+| Customer | Lifetime Value | Visits | Last Visit | Days Absent | Segment |
+|---|---|---|---|---|---|
+| Jordan K. | $631 | 11 | Nov 3, 2025 | 108 | At-Risk VIP |
+| Alex R. | $540 | 9 | Oct 15, 2025 | 127 | At-Risk VIP |
+
+If no customers meet the churn risk threshold: "No high-value customers are currently at churn risk. Great retention!"
+
+### 4. New Customer Acquisition Trend
+
+New customers per month for the trailing 6 months. "New" = first-ever booking at this location.
+
+| Month | New Customers | vs. Prior Month |
+|---|---|---|
+| Sep 2025 | 24 | — |
+| Oct 2025 | 31 | +29% |
+| Nov 2025 | 28 | -10% |
+| Dec 2025 | 42 | +50% |
+| Jan 2026 | 19 | -55% |
+| Feb 2026 | 35 | +84% |
+
+Briefly note any obvious driver of spikes or dips (e.g., "December spike aligns with holiday gift booking surge").
+
+### 5. Loyalty Tier Distribution
+
+Counts of active loyalty accounts by tier, with average points balance:
+
+| Tier | Accounts | Avg Points Balance |
+|---|---|---|
+| Bronze | 183 | 142 |
+| Silver | 64 | 388 |
+| Gold | 22 | 1,204 |
+| Platinum | 8 | 4,210 |
+
+Total enrolled: [sum]. Note percentage of total customer base enrolled in loyalty.
+
+## Privacy Guidelines
+
+- Display customer names as **First name + Last initial** only (e.g., "Sarah M.")
+- Never display full email addresses in output
+- Do not surface booking content details (candle choices, personal notes) in this report — that belongs in guest-letter skill

package/kits/wicks-and-whiskers/skills/demand-forecast/SKILL.md

@@ -0,0 +1,159 @@
+---
+id: demand-forecast
+name: Demand Forecast
+description: Analyzes historical booking and time-slot data to build a demand heatmap, identify peak and quiet windows, and generate concrete staffing recommendations with role-based ratios.
+type: invocable
+version: 1.0.0
+requiresVision: false
+requiresData: true
+---
+
+# Demand Forecast
+
+You analyze {{businessName}} booking patterns to surface when demand peaks, when capacity is underused, and exactly how many staff members should be scheduled for each window. Output is always actionable — a heatmap, peak windows, quiet windows, and a staffing table.
+
+## When to Trigger
+
+Invoke this skill when someone asks:
+
+- "When are we busiest?"
+- "Staffing recommendations for next week"
+- "Should we add a slot on Thursday evenings?"
+- "Capacity planning for Valentine's weekend"
+- "Scheduling question" / "How do I staff this?"
+- Any demand analysis or capacity planning request
+
+## Default Period
+
+If no date range is specified, use the **trailing 90 days** to capture enough historical data for reliable averages. Accept: "last month", "this quarter", "last 6 months", a specific month, or a custom range.
+
+## Aggregate Definition
+
+```typescript
+brain.defineAggregate('bookingsByHourDay', {
+  source: {
+    type: 'Event',
+    where: [
+      { field: 'metadata.entityType', value: 'time-slot' }
+    ]
+  },
+  groupBy: [
+    { field: 'metadata.hour' },
+    { field: 'metadata.date', window: 'day' }   // client-side map to dayOfWeek (0=Sun…6=Sat)
+  ],
+  metrics: {
+    slotCount:   { op: 'COUNT'                                    },
+    totalGuests: { op: 'SUM',    field: 'metadata.reservedCount' },
+    avgGuests:   { op: 'AVG',    field: 'metadata.reservedCount' },
+    stddevGuests:{ op: 'STDDEV', field: 'metadata.reservedCount' },
+    maxGuests:   { op: 'MAX',    field: 'metadata.reservedCount' }
+  }
+});
+
+const heatmapData = await brain.find({
+  aggregate: 'bookingsByHourDay',
+  where: [
+    { field: 'metadata.locationId', value: LOCATION_ID },
+    { field: 'metadata.date', value: { gte: START_DATE, lte: END_DATE } }
+  ]
+});
+// Map groupKey → { dayOfWeek, hour } by parsing the date from each result
+```
+
+After pulling results, map each `groupKey` back to a day-of-week (0–6) and hour (10–22) to build the 7×12 grid.
+
+## Peak and Quiet Window Logic
+
+**Compute global statistics first:**
+- `globalMean` = mean of `avgGuests` across all cells with data
+- `globalStdDev` = STDDEV of `avgGuests` across all cells
+
+**Classification:**
+- **Peak** = cells where `avgGuests > (globalMean + 1 × globalStdDev)`
+- **Hot** = cells where `avgGuests > globalMean` but below peak threshold
+- **Quiet** = cells where `avgGuests < (globalMean - 0.5 × globalStdDev)`
+- **Empty** = cells with no historical slot data (never scheduled)
+
+## Staffing Formula
+
+Apply these role-specific ratios to translate guest counts into staff headcount:
+
+| Role | Ratio | Experience Type |
+|---|---|---|
+| Chandler Guide | 1 per 6 candle bar guests | Kandle Experience, Kitten & Kandle Combo |
+| Klubhouse Kounselor | 1 per 8 kitten lounge guests | Kitten Experience, Kitten & Kandle Combo |
+| POS / Host | 1 always | All sessions |
+
+For a slot with 12 candle bar guests + 12 kitten lounge guests:
+- 2 Chandler Guides + 2 Klubhouse Kounselor + 1 Host = **5 staff minimum**
+
+Round up fractional staff. Never round down — understaffing a kitten lounge is a welfare risk.
+
+## Output Structure
+
+---
+
+### 1. Demand Heatmap
+
+A 12×7 table (hours 10am–9pm as rows, Mon–Sun as columns). Each cell shows average guest count rounded to 1 decimal. Use visual intensity markers:
+
+- 🔴 Peak (> mean + 1σ)
+- 🟡 Hot (> mean)
+- ⚫ (no marker) Normal
+- — Empty/no data
+
+```
+          Mon   Tue   Wed   Thu    Fri    Sat    Sun
+10:00 AM   3.1   2.8   2.4   3.0    4.2    5.8    4.1
+11:00 AM   4.2   3.9   3.7   4.1    6.3   🔴8.2  🟡6.4
+12:00 PM   5.1   4.8   4.4   5.2   🟡7.1  🔴9.4  🟡7.8
+...
+7:00 PM    6.8   5.4   5.9  🟡7.2  🔴9.8  🔴10.4  5.1
+8:00 PM    4.2   3.1   3.4   4.8   🟡6.8  🟡7.2   3.0
+9:00 PM    —     —     —     —      2.4    3.1    —
+```
+
+### 2. Top 5 Peak Windows
+
+Ranked by average guest count:
+
+| Rank | Day | Time | Avg Guests | Max Recorded | Notes |
+|---|---|---|---|---|---|
+| 1 | Saturday | 12:00 PM | 9.4 | 14 | Fully booked 4 of last 8 weeks |
+| 2 | Friday | 7:00 PM | 9.8 | 12 | Valentine's peak in Feb |
+| 3 | Saturday | 7:00 PM | 10.4 | 14 | Consistently sold out |
+| 4 | Friday | 12:00 PM | 7.1 | 10 | Lunch crowd + remote workers |
+| 5 | Sunday | 12:00 PM | 7.8 | 12 | Brunch-adjacent |
+
+### 3. Quietest Windows
+
+Good times to schedule inventory work, kitten care, training, and cleaning without impacting guests:
+
+| Day | Time | Avg Guests | Recommended Use |
+|---|---|---|---|
+| Tuesday | 10:00 AM | 2.8 | Inventory restock, fragrance reorganization |
+| Wednesday | 10:00 AM | 2.4 | Deep clean, litter box maintenance |
+| Thursday | 9:00 PM | — | No bookings — schedule end-of-day kitten care |
+
+### 4. Staffing Recommendation Table
+
+One row per peak or hot window, with minimum staffing:
+
+| Window | Avg Guests | Chandler Guides | Klubhouse Kounselers | Host | Total Staff |
+|---|---|---|---|---|---|
+| Sat 12 PM | 9.4 | 2 | 2 | 1 | 5 |
+| Fri 7 PM | 9.8 | 2 | 2 | 1 | 5 |
+| Sat 7 PM | 10.4 | 2 | 2 | 1 | 5 |
+| Fri 12 PM | 7.1 | 2 | 1 | 1 | 4 |
+| Sun 12 PM | 7.8 | 2 | 1 | 1 | 4 |
+
+### 5. Seasonal Note
+
+If the current month or the next 30 days include a known high-demand period, add a brief callout:
+
+- **February**: Valentine's Day spike. Expect 2–3× normal demand on Feb 14 and the surrounding weekend. Consider adding capacity or a Private buyout option for groups of 4+.
+- **November–December**: Holiday gifting season. Gift card sales surge; book-a-session gift purchases require inventory buffer.
+- **March–May**: Spring surge. New kitten fosters arrive; coordinate schedule with {{rescuePartnerName}}.
+- **Summer (Jun–Aug)**: Family demand increases. Party-of-4+ bookings rise. Private experience marketing performs well.
+
+If no seasonal note applies to the current period, omit this section entirely.