Stop screenshotting shared Gofile folders

Anyone who distributes files through Gofile ends up doing the same job: keeping an inventory of what is actually inside a shared folder. Which files are in there, how big they are, when they landed, how often they get downloaded. The share link answers none of that. It shows one screen, one folder at a time, and the moment you close the tab the information is gone.

So people get creative in the wrong direction. They screenshot the folder view and paste it into a deck. They copy file names into a spreadsheet by hand, size column guessed from the preview. They open the folder every Monday and eyeball the download counter on the two files their boss cares about. It works until the folder has three levels of subfolders, or thirty files, or a colleague asks what changed since last month.

The listing is the data

A Gofile share link looks like a delivery page, but for a public folder it behaves like a catalogue. Every file and subfolder is a row with a name, a size, a type, a checksum, a creation date, and a download counter. That listing is the inventory — it just never leaves the page in a form you can sort, filter, or diff.

The Gofile Scraper API reads that listing for you. You give it public share links (gofile.io/d/...), it returns a flat row per item: name, type, size, MIME type, MD5, timestamps, download count, and the folder page the row came from. Subfolders are followed automatically up to the depth you set, so a top-level link produces the whole tree, not just the first screen.

No account, no password, no session to babysit. The Actor reads public folders only — password-protected content is skipped on purpose, because that content is not public.

Three jobs this actually fits

1. Distribution audits. You publish files through Gofile and want a sheet of what shipped: names, sizes, and the public counter for each file, exportable to JSON or CSV.
2. Change tracking. Scheduled runs snapshot a folder's listing, so a diff between runs shows what appeared, what grew, and which counters moved.
3. Archive inventory. Community managers and researchers who receive shared folders want the tree indexed once — every subfolder, every file, one dataset — instead of clicking through the web view.

If your target is not inventory but the files themselves, walk away: this API reads metadata, not content. Nothing is downloaded, mirrored, or re-hosted. That limit is deliberate — the value is in the structured rows, not in moving bytes.

Run it once before you write a script

Run on Apify.

The flow is the same as any listing job:

1. Paste one or more public folder links.
2. Set how many items and how deep you want.
3. Download the dataset — JSON, CSV, or Excel.

The default input is a small real folder, so you can see the shape of the output in under a minute. Gofile filters heavily automated traffic, so the Actor runs behind residential proxies by default and rate-limits politely rather than hammering the host.

A folder view is not a database

The old cost is not the screenshotting. It is that the inventory silently drifts while nobody is looking: files replaced, counters climbing, a subfolder added that no one noticed. A browser tab cannot diff last week against this week. A dataset can.

The same pattern — a public page that hides structured rows behind a web view — shows up everywhere. We wrote about turning public product pages into price comparisons and monitoring price gaps across geos, two other cases where the source page was never meant to be a spreadsheet but can become one.

If your files are on Gofile and public, the listing can be a dataset. Take the share link, run the Gofile Scraper API, and stop maintaining the folder view by hand.