Load your inventory in about two minutes
No coding, no GitHub account, nothing to install beyond Claude Code, which you already have. You'll run one command, then drop the file it makes into this app. That's the whole thing.
Step 1: Open your terminal
The terminal is a plain window where you type a command and press Enter. Think of it as a text box for talking to your computer. You won't need to know anything about it beyond opening it. Pick your computer below:
- Press ⌘ Space to open Spotlight search.
- Type
Terminaland press Enter. - A small window with a blinking cursor opens. That's it. You're ready.
Step 2: Paste this one command and press Enter
Copy the line below, click into the terminal window, paste it, and press Enter. This one command is the same on every computer: Mac, Windows, or Linux:
npx stack-cleaner@latestIt downloads and runs the tool from npm (the standard place Node tools live), with no curl and nothing piped. The first run may take a few seconds to fetch it.
Prefer not to use npm? Use the per-OS download command
This fetches the same script straight from this site. Make sure your computer is selected, then copy the line, paste it into the terminal, and press Enter.
curl -fsSL https://stackcleaner.com/scan.mjs | nodeWhat it does, plainly: it looks at your own Claude setup and writes a small file called stack-cleaner.json right where the terminal is pointing. It doesn't change anything on your computer, and it doesn't send anything anywhere.
Your API keys and tokens are stripped out before the file is written; they never make it into the file. The script is short, has no dependencies, and you can read it first if you like.
This uses Node, which comes with Claude Code. If you ever see node: command not found (or “not recognized” on Windows), close and reopen the terminal, or install Node from nodejs.org then paste the line again. On older Windows without curl.exe, swap the first part for iwr https://stackcleaner.com/scan.mjs -OutFile stack-cleaner-scan.mjs, or just open /scan.mjs in your browser, save it, and run node on the saved file.
Prefer to look before you run it? Use the two-step version
This saves the script as stack-cleaner-scan.mjs so you can open and read it first, then runs it as a second step. Paste both lines:
curl -fsSL https://stackcleaner.com/scan.mjs -o stack-cleaner-scan.mjs
node stack-cleaner-scan.mjsRunning it creates the data file stack-cleaner.json: that's the one you drop into the app. The result is identical to the one-liner; you just get to inspect the script in between.
Step 3: Find the file it made
When the command finishes, the terminal prints a line like ✓ wrote stack-cleaner.json with the exact location. The file is named stack-cleaner.json and it's saved in whatever folder the terminal was pointing at: for a terminal you just opened, that's your home folder (the one with your name on it). On a Mac that's /Users/you; on Windows it's C:\Users\you (the first entry under Home / Quick access in File Explorer). Trust the printed ✓ path over guessing.
Don't worry about hunting for it on disk: in the next step you can open your file browser to that folder and pick it (on a Mac you can also drag it straight out of the terminal).
Step 4: Drop it into the app
Open the inventory page, then drag the stack-cleaner.json file onto it, or click Choose file and pick it. From that moment on, everything stays inside your browser; nothing is uploaded.
Your whole setup, finally legible
The scan script and full source live on GitHub. Clone the repo and run node public/scan.mjs, or install it once with npm i -g stack-cleaner and run stack-cleaner.
You don't have to scan anything to explore. Open the demo inventory and click around with sample data.