v0.2.0a1 — Menu sales and food-cost analysis

Menu sales and food-cost review through a local CLI and MCP-compatible assistants. This release provides contribution rankings, weighted food-cost percentages, and a documented workflow for reviewing menu exports.

Included

• Local CSV report and JSON output via restaurant-ops.
• Item and menu analysis through the official MCP Python SDK v2.
• Clear validation errors, loss-making-item warnings, and no-sales handling.
• Synthetic example menu, weekly review guide, operator pilot checklist, and contribution templates.

Engineering and verification

46 tests passed against the packaged wheel in a fresh Python 3.12 environment, including command-line use and a real stdio MCP connection. Both wheel and source distribution passed package metadata checks. CI covers Python 3.10–3.14, the minimum SDK version, and package installation.

Try it

Download and extract the source archive below. In its folder, create and activate a Python 3.10+ virtual environment, run python -m pip install ., then restaurant-ops examples/menu.csv. See the README for Windows activation and MCP client setup.

The included synthetic example produces 8,152.50 in sales and 5,702.00 in contribution after ingredients across 615 units.

Calculation scope

Contribution is sales minus recipe ingredient costs. Labor, rent, fees, waste, and inventory reconciliation are outside this calculation. Sales equal supplied price times supplied units; recipe costs represent expected per-portion usage. Returns, zero-price comps, weighted quantities, live POS imports, period comparisons, and price-change scenarios are not supported. Use one currency per file.

For periods with no sales, the text report shows food-cost percentage as N/A; JSON retains 0.0 with a warning. Displayed amounts are rounded to two decimal places.

Install from the GitHub source archive; this release is not published to PyPI.

Release documentation