prep-listlisted

# Prep List You turn a day's forecast into two working documents — a station-by-station prep list and one consolidated order — using a bundled script so the aggregation is exact. ## Method — run the script, never total ingredients by hand 1. **Collect** each dish on today's service: its **forecast covers**, its **station**, and its **per-portion ingredients** (name, quantity, unit). Ask for covers if the user only gives the menu. 2. **Build the input JSON:** ```json { "service": "Fri dinner", "dishes": [ {"name": "Beef Burger", "covers": 60, "station": "grill", "ingredients": [ {"name": "Beef patty", "qty": 1, "unit": "each"}, {"name": "Cheddar", "qty": 30, "unit": "g"} ]} ] } ``` Quantities are **per portion** — the script multiplies by covers. Sample: `${CLAUDE_SKILL_DIR}/scripts/example_service.json`. 3. **Run it:** ```bash python3 ${CLAUDE_SKILL_DIR}/scripts/prep_list.py /tmp/service.json ``` Use the file-argument form so it matches the pre-approved `allowed-tools` rule. It returns a PREP list grouped by station and an ORDER that sums every ingredient across all dishes (g/kg and ml/L combined). **Trust its totals.** 4. **Then add the judgment.** Round order quantities up to purchase/pack units, add a waste or safety buffer if the user wants one, and note what must be prepped fresh vs. ahead. ## Guardrails - Always run `${CLAUDE_SKILL_DIR}/scripts/prep_list