thinkhard.SI

Prompt repository Operations

Decide where stock goes in a warehouse so picking is faster

Pickers walk too far or pick the wrong items, and you want to move stock so the fastest movers sit closest to dispatch.

Registry no.
OPS-050
Collection
Operations
Laid out for
GPT
Score
10 out of 10
Blanks to fill
9

OperationsGPT

The prompt

Score 10 of 10, Excellent

Use it when Pickers walk too far or pick the wrong items, and you want to move stock so the fastest movers sit closest to dispatch.

  • {{audience}}
  • {{move_window}}
  • {{period}}
  • {{pick_data}}
  • {{item_data}}
  • {{locations}}
  • {{constraints}}
  • {{order_data}}
  • {{max_moves}}
Full prompt and why it works
Prompt
# Role
You help warehouse managers decide where each item should live so that pickers walk less and pick fewer wrong items.

# Goal
Write a slotting plan from the data below, for {{audience}} to approve and carry out over {{move_window}}, so that the fastest-moving items sit closest to dispatch at the easiest height, and every move is worth the labor it takes.

# Material
Pick data by item for {{period}} (item, description, pick lines, units picked, current location):
"""
{{pick_data}}
"""

Item sizes and weights, with case or unit dimensions:
"""
{{item_data}}
"""

Locations: types, sizes, weight limits, and distance or zone rank from dispatch: {{locations}}
Items that must stay put or be kept apart (hazardous, chilled, high value, fixed locations): {{constraints}}
Order data, if we have it, showing items often picked together: {{order_data}}

# Output
1. Slotting rules: up to eight short rules for this warehouse, drawn from the data, for example "Items in the top 10% of pick lines go at waist to shoulder height in the aisles nearest dispatch; items over [weight] go on the lowest level." That example shows the form only.
2. Velocity classes: a table of items grouped by pick lines (not units), with the cutoffs used and the share of all picks each group accounts for.
3. Moves: a table with the columns Item, Current location, Proposed location, Pick lines, Reason, Sequence. Limit it to the {{max_moves}} moves with the most effect, and order them so no location is needed before it is emptied.
4. Checks before moving: the new location holds at least one full replenishment, weight limits are met, and items that look alike are placed apart to cut wrong picks.

# Constraints
- Rank by pick lines, because a picker walks to a location once per line, whatever the quantity.
- Use only the locations and sizes given; do not invent capacities or distances.
- Do not move items listed as fixed, and do not break any separation rule.
- A good plan is one that saves most of the walking with the fewest moves; say which moves matter most.
- Keep the written part under 300 words, not counting the tables.

# When data is missing
If an item has no size or weight, or a location has no capacity, leave it out of the moves, list it under "Data needed", and do not guess.

Variables

{{audience}}
Who approves and does the moves, for example "the warehouse manager and the inventory team".
{{move_window}}
When the moves can happen, for example "two Saturday shifts".
{{period}}
The period the pick data covers, for example "the last 13 weeks".
{{pick_data}}
One row per item: item code, description, pick lines, units picked, current location, exported from your warehouse system.
{{item_data}}
Item dimensions and weights, per unit or per case.
{{locations}}
Location types (pallet, shelf, carton flow), sizes, weight limits, and how far each zone is from dispatch.
{{constraints}}
Items that must stay where they are or be kept apart, and why, or "none".
{{order_data}}
Pairs or groups of items often on the same order, or "none".
{{max_moves}}
The most moves the team can do in the window, for example "50".

Why it works

  • What good looks like. Items are ranked by pick lines rather than units, because walking is driven by visits to a location, and the plan is judged on the most walking saved from the fewest moves.
  • Length. A cap on moves sized to the window, in an order that never needs a location before it is free, keeps the plan doable over a weekend.
  • If unsure. Items or locations with missing sizes stay out of the moves and go on a Data needed list rather than being slotted by guesswork.

How to use it

  1. Press Fill in blanks and type your details into each blank. What you type stays in this browser.
  2. Copy the filled prompt into the AI tool you already use, such as ChatGPT, Claude or Gemini, or open it in the Grader or the Prompt Tester here.
  3. Before you rely on it, try it on three or four real examples. The Prompt Tester runs them all at once and shows which ones go wrong.
Use it from code

Get this prompt as JSON from the free REST API, with your own text in the blanks, or let an AI assistant fetch it through the MCP server. No key.

curl -s https://thinkhard.si/api/v1/prompts/warehouse-slotting-plan

With the MCP server added to Claude, ChatGPT, Cursor or VS Code, call get_prompt with {"prompt": "warehouse-slotting-plan"}, or just ask for it by name.