Menu Costing
**an alternative to MarketMan's recipe costing workflow included with your subscription.** Generate recipe cost cards, food cost analysis, menu engineering matrix (Stars/Plowhorses/Puzzles/Dogs), and pricing recommendations — all loc
Plugin ID
pf-menu-costing
Category
food
Version
v1.0
Downloads
pf-menu-costing v1.1
An alternative to MarketMan's recipe costing workflow, included with your subscription. Generate recipe cost cards, food cost analysis, menu engineering matrix (Stars/Plowhorses/Puzzles/Dogs), and pricing recommendations — all locally, no subscription, no contract. Now with professionally structured domain knowledge: NRA benchmarks, Kasavana-Smith menu engineering framework, USDA yield factors, and pricing psychology.
Installation
- Download the
pf-menu-costing.pluginfile - Open Claude Desktop and navigate to Settings > Plugins
- Click Install Plugin and select the downloaded
.pluginfile - The plugin will be installed and available immediately
Note: All data stays local on your machine. No external API calls or cloud storage required.
Why This Exists
Restaurant operators pay $1,800–$4,800/year for MarketMan to answer two questions: "What does each dish cost me?" and "Which items should I promote or cut?" This plugin gives you those answers in 15 minutes using a spreadsheet and Claude — included with your subscription.
Quick Start
- Install the plugin and run
/menu-costing:menu-costing-setup - Fill in
inbound/ingredient-prices-template.csvwith your actual ingredient costs - Run
/menu-costing:menu-costing-import-pricesto load the price database - Run
/menu-costing:menu-costing-cost-cardand enter your recipes (conversationally or via CSV) - Run
/menu-costing:menu-costing-fullto complete the full analysis — cost cards, food cost report, menu matrix, and pricing recommendations are waiting inoutbound/
Commands
| Command | Description |
|---|---|
/menu-costing:menu-costing-setup |
Initialize workspace, install dependencies, create config |
/menu-costing:menu-costing-import-prices |
Import ingredient prices from CSV/XLSX |
/menu-costing:menu-costing-cost-card |
Generate recipe cost cards |
/menu-costing:menu-costing-food-cost |
Analyze food cost % across all recipes |
/menu-costing:menu-costing-menu-matrix |
Build the Stars/Plowhorses/Puzzles/Dogs matrix |
/menu-costing:menu-costing-pricing-recs |
Get pricing recommendations |
/menu-costing:menu-costing-full |
Run all stages end-to-end (with resume support) |
/menu-costing:menu-costing-status |
Show pipeline progress |
How It Works
inbound/ingredient-prices.csv
↓ [import-prices]
.menu-costing/prices.json
↓ [cost-card]
processing/menu-costing-cost-cards.json → outbound/recipe-cost-cards.docx
↓ [food-cost]
processing/menu-costing-food-cost.json → outbound/food-cost-analysis.docx
↓ [menu-matrix] + sales volume
processing/menu-costing-matrix.json → outbound/menu-engineering-matrix.xlsx
↓ [pricing-recs]
outbound/pricing-recommendations.docx + RESULTS.md
All outputs land in outbound/. Open them in Word or Excel — no login required.
What It's an Alternative To
| SaaS Tool | Annual Cost (SMB) | Key Overlap |
|---|---|---|
| MarketMan | $1,800–$4,800/yr | Recipe costing, food cost %, menu engineering |
| BlueCart | $2,400–$6,000/yr | Ingredient cost tracking, recipe analysis |
| Craftable | $1,200–$3,600/yr | Menu engineering, cost cards |
Key Differentiators
Beyond the Included with your subscription, pf-menu-costing offers advantages MarketMan can't match:
- Your data stays local — ingredient costs and recipe formulas is processed within the Cowork environment
- No lock-in — all outputs are standard DOCX and XLSX files you can open anywhere
- Fully customizable — edit the document templates to match your brand
- No training required — just tell Claude your recipes in plain English
- Weekly price updates take 2 minutes — update one CSV, re-run, done
Feature Comparison
| Feature | pf-menu-costing | MarketMan | BlueCart |
|---|---|---|---|
| Recipe cost cards | ✅ Full | ✅ Full | ✅ Full |
| Food cost % analysis | ✅ Full | ✅ Full | ✅ Full |
| Menu engineering matrix | ✅ Full (Stars/Plowhorses/Puzzles/Dogs) | ✅ Full | ⚠️ Partial |
| Pricing recommendations | ✅ Full (AI-generated) | ⚠️ Partial | ❌ None |
| Invoice scanning | ❌ Manual CSV import | ✅ Full | ✅ Full |
| POS integration (live sales) | ❌ Manual CSV | ✅ Full | ✅ Full |
| Real-time price alerts | ❌ Weekly manual re-import | ✅ Full | ✅ Full |
| Multi-location | ❌ One folder per location | ✅ Full | ✅ Full |
| Data privacy | ✅ Local only | ❌ Cloud-stored | ❌ Cloud-stored |
| Customizable templates | ✅ Edit in Word | ❌ Fixed format | ❌ Fixed format |
| Monthly cost | included with your subscription | $150–$400 | $200–$500 |
Estimated Cost per Use
Disclaimer: Token estimates are approximate and based on typical usage patterns measured from skill prompt sizes. Actual costs vary with input data size, conversation length, and complexity. Estimates use Claude Sonnet 4.6 pricing ($3/1M input, $15/1M output). Cowork and Claude Desktop subscription users (Pro/Max/Team) are not charged per-token — these estimates apply only to direct Anthropic API usage. Running stages individually in fresh sessions uses fewer input tokens than running the full pipeline sequentially, because pipeline mode accumulates conversation history across stages.
Per skill (run individually in a fresh session):
| Stage | Skill Prompt | User Input | Total Input | Output | Est. Cost |
|---|---|---|---|---|---|
| menu-costing-menu-matrix | ~8.5K | ~800 | ~12.2K | ~6.0K | ~$0.13 |
| menu-costing-food-cost | ~7.6K | ~800 | ~11.1K | ~6.0K | ~$0.12 |
| menu-costing-import-prices | ~4.1K | ~2.0K | ~8.9K | ~2.0K | ~$0.06 |
| menu-costing-cost-card | ~6.1K | ~800 | ~9.7K | ~6.0K | ~$0.12 |
| menu-costing-pricing-recs | ~7.9K | ~800 | ~11.5K | ~6.0K | ~$0.12 |
| Standalone total | ~53.3K | ~26.0K | ~$0.55 |
Full pipeline (all stages in one session — context accumulates):
| Stage | Base Input | + History | Total Input | Output | Est. Cost |
|---|---|---|---|---|---|
| menu-costing-menu-matrix | ~12.8K | 0 | ~12.8K | ~6.0K | ~$0.13 |
| menu-costing-food-cost | ~11.8K | ~6.8K | ~18.6K | ~6.0K | ~$0.15 |
| menu-costing-import-prices | ~9.5K | ~13.6K | ~23.1K | ~2.0K | ~$0.10 |
| menu-costing-cost-card | ~10.3K | ~17.6K | ~27.9K | ~6.0K | ~$0.17 |
| menu-costing-pricing-recs | ~12.1K | ~24.4K | ~36.5K | ~6.0K | ~$0.20 |
| Pipeline total | ~119.0K | ~26.0K | ~$0.75 |
Running the full pipeline once typically costs $0.52–$0.97 in API tokens (Claude Sonnet 4.6).
AI-Powered Features (12 features)
- NRA Benchmark Calibration: Food cost targets calibrated to restaurant segment (QSR 28%, casual dining 32%, fine dining 35%, steakhouse 38%) per NRA Restaurant Industry Operations Report
- Kasavana-Smith Menu Engineering: Academic-grade quadrant classification (Stars/Plowhorses/Puzzles/Dogs) based on the original 1982 Michigan State University framework with median-based thresholds
- USDA Yield Factor Awareness: AP-to-EP cost conversion using USDA Agriculture Handbook No. 102 yield data for proteins, produce, and dairy
- Food Cost Percentage Analysis: Computes overall and per-recipe food cost % against segment-appropriate thresholds
- Category-Based Aggregation: Groups recipes by menu category with weighted cost averages compared to NRA category benchmarks (soups 12–20%, entrées 28–38%)
- Cost Threshold Flagging: Highlights items above warning and critical thresholds calibrated to restaurant type
- Cost Optimization Recommendations: AI recommendations grounded in NRA benchmarks and professional restaurant consulting methodology
- Pricing Psychology Engine: Recommendations informed by charm pricing, price anchoring, left-digit effect, and menu position psychology (Cornell Hospitality Quarterly research)
- Contribution Margin Analysis: Both FC% and absolute CM metrics with guidance on when each matters (NRA: high-volume → FC%, fine dining → CM)
- Seasonal Price Intelligence: Flags seasonal price spikes during import to prevent unnecessary menu price changes
- Prime Cost Context: Frames food cost within the full prime cost equation (food + beverage + labor ≤ 65% of revenue)
- Strategic Price Increase Sequencing: Recommends which items to reprice first by quadrant (Plowhorses → Puzzles → Stars → Dogs) with 3–5% annual tolerance guidance
Known Limitations & Workarounds
This is a Claude Cowork plugin, not a traditional SaaS application. It has architectural constraints you should know about before adopting it.
| Limitation | Impact | Workaround |
|---|---|---|
| No invoice scanning | Prices must be entered manually via CSV | Update ingredient-prices-template.csv weekly from your invoices — takes 5–10 min |
| No POS integration | Sales volume for menu matrix must be entered manually | Export weekly sales report from your POS as CSV and upload to inbound/ |
| No real-time price alerts | Price spikes between weekly imports go unnoticed | Import prices weekly; check price-delta.json after each import for changes |
| No multi-user collaboration | Single-user, session-based | Use one workspace per location; share the outbound/ DOCX/XLSX files with your team |
| Session-based operation | Analysis only runs when you're in a Cowork session | No background monitoring; run weekly as a routine |
| No e-signature integration | Generated reports are for internal use only | Export to PDF from Word before sharing externally |
Context & Performance Guide
Session Management
Claude's context window (200K tokens) is shared across the entire session. For best results:
- Start a fresh Claude session before running
/menu-costing:menu-costing-full - If you've been chatting extensively in the current session, open a new one for cleaner output
- Each stage can also be run independently — the plugin saves state between stages
Data Volume Guidance
Tested with up to 150 recipes and 300 ingredients in a single run. For typical restaurant use (40–80 recipes, 100–200 ingredients), performance is excellent. If you have an unusually large menu:
- Run stages individually instead of using
/menu-costing:menu-costing-full - Split your menu into sections (Lunch/Dinner/Brunch) and run separate analyses
Tips for Best Results
- Provide ingredient prices as a CSV file rather than entering them conversationally — file reading is more context-efficient
- Keep recipe names consistent between your price list and recipe input — matching is case-insensitive but spelling matters
- Enter recipes in batches (10–20 at a time) for the cleanest output
- If output quality degrades, start a new session and re-run the current stage — the plugin saves progress
Degradation Signals
If responses become shorter, skip sections, or produce incomplete tables, start a new session and re-run the current stage from the pipeline.
Requirements
- Claude Desktop or Cowork mode
- Python 3 (pre-installed in Cowork VM)
- python-docx and openpyxl (auto-installed by setup)
- Ingredient prices in CSV or Excel format
- Recipes (enter conversationally or as CSV)
Important Disclaimers
- AI-Generated Content: This plugin uses AI (LLM) technology which can produce inaccurate or incomplete outputs. All content should be treated as a starting point and reviewed for accuracy before use.
- Not Professional Advice: Outputs do not constitute legal, financial, tax, medical, or other professional advice. Consult qualified professionals before making decisions based on generated content.
- No Compliance Guarantee: References to industry standards, regulations, or guidelines are for informational purposes only. This plugin does not guarantee compliance with any law or regulation. Users are responsible for verifying all outputs meet their specific regulatory requirements.
- No Endorsement or Affiliation: Mention of third-party products, standards, or organizations does not imply endorsement, partnership, or certification by those entities.
- Not Financial or Tax Advice: Financial calculations, tax estimates, and related outputs are for informational purposes only. Consult a qualified CPA or financial advisor for tax and financial decisions.
Ready to use Menu Costing?
Download this free plugin and start using it in Claude today.
Need something different?
We build custom plugins tailored to your exact workflow.