Forward-Pick Replenishment Trigger

The problem this kills

A picker walks up to a forward-pick face and it's empty. The whole wave stalls. Someone radios for a replen, an associate hunts the reserve location, and meanwhile the order sits. Multiply that by every fast-mover during a peak push and you've got missed cutoffs and frustrated people.

The fix isn't more radios - it's looking ahead. Your pick-face minimums, the stock actually sitting on the face right now, and the demand from today's open orders already tell you which faces are going to run dry. Nobody has time to reconcile three spreadsheets between waves, so it doesn't happen, and the faces keep going empty.

This tool does that reconciliation for you, every cycle, and hands your supervisor a clean, approved move list before the face ever hits zero.

What you'll build

A small web app your replenishment team logs into that:

• Pulls in your pick-face min/max and on-hand, your reserve/bulk on-hand, and your open-order demand (from a CSV or Google Sheet - no integration required).
• Calculates, for every face, whether on-hand minus committed demand will drop below the minimum.
• Drafts replenishment moves from the right reserve location to the forward face, sized to top the face back up toward max.
• Prioritizes the faces feeding today's waves first.
• Routes the proposed move list to a supervisor, who reviews and approves before anything is released to the floor.
• Exports the approved moves as a CSV in the exact columns your WMS or floor team expects, and shifts the on-hand to reflect the move.

What's inside the Implementation Plan

A complete, paste-and-go runbook for Claude Code. The very first thing it does is interview you about your warehouse - your face/location naming, how minimums are set, where demand lives, your wave schedule, your peak volumes, and your messy exceptions - then it reflects a short tailored spec back to you for a thumbs-up before it builds a single thing. You get a tool shaped around your operation, not a generic template.

Inside:

• The discovery interview (8-12 warehouse-specific questions) and how the agent turns your answers into the data model.
• Step-by-step build prompts you copy and paste - data import, the replenishment calculation, the supervisor approval gate, the CSV export, and the on-hand shift.
• Built-in governance (login, row-level security, audit trail, approval gate, duplicate guards) wired in from the start.
• A "No API yet?" fallback so the whole thing works from spreadsheets today.

The governance it includes (this is the point)

Replenishment touches your real inventory, so this isn't a toy:

• Login so only your team can use it.
• Row-level security so each site or org only ever sees its own faces and stock.
• A full audit trail - who generated a move list, who approved it, and exactly when.
• A hard human-in-the-loop gate - the AI drafts the moves, a supervisor reviews and approves, and only then is the list released and the on-hand shifted. Nothing commits to your system of record automatically.
• Duplicate guards - the same face can't be queued twice in one replen cycle, so you never send two associates to the same location.

Who it's for

Replenishment associates, pick/pack supervisors, and inventory control staff who live in spreadsheets and the WMS and want to get ahead of empty faces - without writing code or waiting on IT.

You've got this. Paste the first prompt and let the interview tailor it to your warehouse.