Overstock & Dead-Stock Flagger

The problem this kills

Somewhere in your warehouse, cash is quietly sitting on a shelf. SKUs you over-ordered last season. Items that haven't moved in 200 days. Stuff nobody flags until the year-end count, when finance asks why inventory value is up and turns are down.

The usual "system" for catching this is a planner squinting at a giant spreadsheet once a quarter, sorting by quantity, and eyeballing what looks fishy. Slow, inconsistent, and easy to skip when things get busy. Truly dead items get reordered again on autopilot. Seasonal-but-valid stock gets liquidated by mistake. And nobody has a defensible, dated record of who decided to mark something down.

This tool turns that quarterly guessing game into a repeatable review: import your numbers, let the math flag the overstock and the no-movers, see the dollars each one is tying up, and walk a manager through approve/reject decisions before a single reorder block or markdown leaves the building.

What you'll build

A small, private web app for your team that:

• Imports your on-hand + cost and your sales/usage history from a CSV or Google Sheet (no integration required to start).
• Flags two kinds of trouble: overstock (months-of-supply far above your threshold) and dead stock (no movement for N days), while separating truly dead items from seasonal-but-valid ones.
• Estimates the tied-up cash for each flagged SKU, weighted by carrying cost, so you triage the biggest dollars first.
• Recommends an action per SKU — stop reorder, markdown, liquidate, or keep — that a person can override.
• Puts a manager in the loop: nothing is exported until a planner or manager reviews and approves each recommendation.
• Exports a clean action list as CSV in the exact columns your ERP or purchasing system expects.

What's inside the Implementation Plan

• It starts by interviewing you about your business. Before writing any code, the plan has the AI agent ask about your current review process, your ERP or spreadsheet, your exact field names and SKU conventions, your typical and peak volumes, your overstock and dead-stock thresholds, and your messy edge cases (seasonal lines, kits/bundles, new items with no history). It reflects a short tailored spec back to you and waits for your thumbs-up — so you get a tool shaped around your inventory, not a generic template.
• A copy-paste prompt for every build step, written for non-coders.
• The full data model, flagging math, and dollar-impact logic, adapted to your answers.
• The manager approval screen and the audit trail behind it.
• The CSV import and the CSV export in your system's column layout.
• A "How to know it works" verification checklist.

The governance it includes (this is the point)

This isn't a throwaway script — it's an internal tool you can stand behind in a finance review:

• Login so only your team can open it.
• Row-level security so each organization only ever sees its own data.
• A complete audit trail — who flagged, who approved or overrode, what action, and exactly when.
• A hard human-in-the-loop approval gate — the AI drafts the recommendations, a manager reviews and approves, and only then is the action list committed and exported.
• Duplicate guards so the same SKU can't be processed twice in the same review cycle.

Who it's for

Inventory and demand planners, finance and FP&A partners, and category managers who own reorder and markdown decisions — anyone who needs to free up cash from slow inventory without accidentally killing a seasonal winner.

You've got this — paste the first prompt and let the interview tailor the rest.