Stashcard gives your Steam inventory a receipt with wallet and cash prices kept separate
The free Windows tool exports inventory estimates and their provenance. Its README also contradicts itself about platform support.

A Steam inventory can look impressively valuable until someone asks what that number means.
Stashcard’s terminal workflow puts the market, time window and estimated fees beside the total, then prints the result as a shareable card. The project says, “Nothing unpriced is ever counted as zero.” Missing data gets called out, along with fallback sources. The exported PNG has a JSON sidecar for checking the numbers or rendering it again.
The Windows portable download is the clearest advertised route. The README elsewhere claims Linux and macOS work, but its installation warning says the packaged companion runtime is Windows-only. That contradiction remains unresolved; the badges cannot settle it.
The difference between a listed price and a completed sale runs through the tool’s output. LOWEST means the current floor listing, while AVG is a mean of actual sales over a chosen window. A seller can ask an impressive amount for an item that nobody buys; the second number helps put the first in context.
That makes the time window worth keeping with the image. A short burst of sales and a longer history can yield different averages without either calculation being wrong. Stashcard offers several windows, and its README says numbers retain their venue and window. A screenshot with those labels cropped away would remove information the software deliberately preserved.
The exported JSON is also useful beyond checking arithmetic. The documented render command can make another image from a saved sidecar without repeating a live lookup. That separates presentation changes from a fresh market scan, so changing a theme need not quietly change the estimate underneath it.
The project includes an offline demo built around a fictional profile and dated prices. That is a sensible first stop for seeing its receipt, inventory grid and marketplace layouts. A demo proves the format can be produced; a live scan still has to find the person’s actual inventory and available prices.
The README offers history-based comparisons as well. For two saved valuations to mean much, their market and window choices need to match. Otherwise an apparent change in the stash could come from changing the measuring method. Store credit and cash already differ; mixing yesterday’s sales average with today’s cheapest listing adds another difference to untangle.
Someone tempted to treat the PNG as a sale offer should pause at the fee and coverage details. The useful result is an estimate with visible assumptions. Whether a specific buyer will pay that amount remains outside the image.
The free MIT-licensed project offers an unusually inspectable alternative to a mysterious valuation box. Its next useful improvement would be a consistent platform-support statement that agrees with the shipped runtime.