Runwaystock

io.github.GillMacedoPublished by GitHub account GillMacedov0.2.0Updated Nov 3, 2026

Find what to reorder, stockout dates and dead stock from sales and stock CSVs, on your machine.

VerifiedSTDIODesktop onlyData & AnalyticsBusiness & Commerce

Overview

AI-generated overview

Local MCP server that reads sales and stock CSVs to compute reorder points, order quantities, stockout dates and dead stock per SKU.

What you can do
  • Run an inventory check on a sales CSV and a stock CSV, returning each SKU's status and suggested order quantity, most urgent first.
  • Inspect an unfamiliar CSV to see which columns are detected and how many rows can be read, without returning row contents.
  • Statuses cover order now, order soon, ok, overstocked and no recent sales, with reorder points, safety stock and stockout dates.
  • Both tools are read-only and never place orders or use the network.
When to use it
Useful for e-commerce stores, retailers and small businesses that plan inventory in spreadsheets and want an assistant to say what to reorder, how much, which products run out first, and how much money sits in dead stock.
Requirements
Runs locally over stdio; Node.js is needed to launch it through npx, or Go 1.27 or newer to build the binary. No account, API key or network access is required. It needs a sales CSV and a stock CSV, with at least 28 days of sales history for reliable figures.
Before you install
The check tool returns SKUs and quantities to the MCP client, so they reach whichever AI model you use; the CLI keeps everything local. Results are estimates: the model ignores seasonality, promotions and lost sales, and it has only been tested on made-up data, so verify before buying. It reads CSV only, up to 50 MB.

Installation

In SourceWeft

  1. Open Runwaystock in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.

Other MCP clients

Follow the launch instructions in the repository.

README

runwaystock: inventory reorder points from your sales and stock files

[npm] [CI] [MIT license] [MCP Registry]

Know what to reorder, how much to buy, when each product runs out, and how much money is stuck in dead stock. For e-commerce stores, retailers and small businesses that plan inventory in spreadsheets.

Inventory planning software is expensive, and spreadsheet formulas break. runwaystock reads a sales export and a stock count as CSV files, and applies the standard reorder point and safety stock formulas to every SKU. It's an MCP server and a single Go binary, with no account, no API key and no network access.

What you can ask your agent

  • "What do I need to reorder this week, and how much?"
  • "Which products will run out first?"
  • "How much money is tied up in dead stock?"
  • "Which products am I overstocked on?"

Install

As a Claude Code plugin, which adds the MCP server and a skill that tells Claude when to use it. You need Node, which runs the server through npx:

claude plugin marketplace add GillMacedo/runwaystockclaude plugin install runwaystock@runwaystock-marketplace

For other MCP clients:

json
{  "mcpServers": {    "runwaystock": {      "command": "npx",      "args": ["-y", "runwaystock", "serve"]    }  }}

Or build the binary with Go 1.27 or newer:

go install github.com/GillMacedo/runwaystock@latest

Usage

Without installing anything, run npx runwaystock check sales.csv stock.csv. This example uses the made-up files in examples/:

$ go run . check examples/sales.csv examples/stock.csvAs of 2026-09-30, using 90 days of salesORDER NOW        WIDGET-RED: buy 93  costs 418.50  (40 on hand, 2.83 a day, out by 2026-10-14, reorder at 48)OK               MUG-LARGE  (70 on hand, 35 days left)OK               CAP-DEMO  (300 on hand, 51 days left)OVERSTOCKED      WIDGET-BLUE  (120 on hand, 277 days left)OVERSTOCKED      BAG-CANVAS  (500 on hand, 3462 days left)NO RECENT SALES  POSTER-OLD  (60 on hand, worth 120.00)NO RECENT SALES  TSHIRT-UNSOLD  (35 on hand, worth 420.00)NOT IN STOCK FILE  STICKER-NEW sold recentlySKIPPED sales row 3281 to order now, costing 418.50. 540.00 tied up in stock with no recent sales.

Set your own assumptions with --lead 21, --cover 45, --window 60 and --service 99. A lead time in the stock file overrides --lead for that SKU. Add --json for machine-readable output.

How it calculates

safety stock  = z x standard deviation of daily sales x square root of lead daysreorder point = daily sales x lead days + safety stockorder qty     = daily sales x (lead days + cover days) + safety stock - on hand

Daily sales are averaged over calendar days, including days with no sales. Returns reduce demand. Orders round up to any minimum order quantity (MOQ) in the stock file. Service levels of 90, 95 and 99 use z values of 1.28, 1.65 and 2.33.

StatusMeaning
Order nowOn hand is at or below the reorder point
Order soonIt will reach the reorder point within 7 days
OkEnough stock for now
OverstockedMore than 180 days of stock left
No recent salesNo net sales in the window

MCP tools

ToolWhat it does
inventory_checkReturns each SKU's status and order, most urgent first
inventory_inspect_csvShows detected columns and row counts, without row contents

Both are read-only, and neither can place orders. inventory_check returns SKUs and quantities to the MCP client, so they reach whichever AI model you use. The CLI keeps everything on your machine.

FAQ

How do I calculate a reorder point? Use the formula above, or run npx runwaystock check sales.csv stock.csv to do it for every SKU.

Does it work with Shopify, Amazon or WooCommerce? It reads CSV and recognizes common column names such as Lineitem sku and Lineitem quantity, and you can name others with --sales-columns. It doesn't connect to those platforms, and it hasn't been tested on each one's real exports yet.

How much sales history do I need? At least 28 days, or it warns that figures are rough. It uses the last 90 days by default.

Limits

  • It assumes recent sales predict the near future, so it doesn't see seasonality, promotions or growth.
  • Sales lost while out of stock look like low demand, which understates what you'd need.
  • The safety stock formula assumes fairly regular demand and a fixed lead time, so treat results as a starting point and check before buying.
  • It reads CSV only, up to 50 MB, and doesn't handle bundles or unit conversions.
  • So far it has been tested on made-up data, not real exports. If yours fails, please open an issue with the column headers, not your data.

Contributing

Contributions are welcome, from people and from people working with AI agents. See CONTRIBUTING.md. Good first issues:

  • Add column names from more e-commerce platforms and languages
  • Add --out to write a purchase list to a CSV
  • Group the order list by supplier
  • Add a seasonality option that compares with the same period last year
  • Add a plugin icon

AI disclosure

This project was largely generated with Claude Code and reviewed by its maintainer.

About me

I'm Giulliano Macedo, and I build developer tools at GoTech Solutions. Follow along or get in touch:

If runwaystock saved you from a stockout, a star helps others find it.

MIT licensed.

Built by Giulliano Macedo, GoTech Solutions · LinkedIn · X

Source: README.md at commit 3e91907

Tools

2

inventory_check

Check stock and reorder points

Works out what to reorder and how much from a sales CSV and a stock CSV, with reorder points, safety stock, stockout dates and dead stock. Use for inventory planning, purchase orders, or finding slow-moving stock. Read-only, and never uses the network. Results include SKUs and quantities from the files, so they reach the model. It can't place orders.

Parameterslimitlead_dayscover_dayssales_pathrequiredstock_pathrequireddate_formatwindow_dayssales_columnsservice_levelstock_columns

inventory_inspect_csv

Inspect a sales or stock CSV

Shows which columns runwaystock detects in a CSV and how many rows it can read, without returning row contents. Use if a file is unfamiliar or a check failed.

Parameterskindrequiredpathrequired

Version history

2
  1. v0.2.0LatestNov 1, 2026
  2. v0.1.0Oct 31, 2026

Score

90% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 days25 pts
  • Auth declared10 pts

    Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.

  • Tools listedRequired15 pts
  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts