[Feature Request] No-op Output Mode

[tldev]

Hi,

I have a few automations in the background for when downloads complete that run into race conditions with Shelfmark. I would prefer Shelfmark do nothing with the download and I handle the movement. Can you provide a no-op output mode where the download is just left where it is?

[mrskizzex]

I agree, this is very much needed.

This would also help with the issue where I have set download location for qbittorent to /qbt-audiobook-download in both qbt and shelfmark and after qbt moves the files from temp download location to /qbt-audiobook-download/, shelfmark moves them again but without a folder? I'm unsure why is that desirable.

2026-04-10 16:07:38,016 - shelfmark.download.orchestrator - INFO - orchestrator.py:374 - Task 18ceef75dbfb25ce9c3c97cb1bbe4cb8: download finished; starting post-processing
2026-04-10 16:07:38,021 - shelfmark.download.postprocess.router - INFO - router.py:47 - Task 18ceef75dbfb25ce9c3c97cb1bbe4cb8: using output mode folder
2026-04-10 16:07:39,821 - shelfmark.download.outputs.folder - INFO - folder.py:201 - Task 18ceef75dbfb25ce9c3c97cb1bbe4cb8: transferred 10 file(s) to /qbt-audiobook-download (ops: copy=10)
2026-04-10 16:07:39,849 - shelfmark.download.orchestrator - INFO - orchestrator.py:391 - Task 18ceef75dbfb25ce9c3c97cb1bbe4cb8: post-processing complete
2026-04-10 16:07:39,856 - shelfmark.download.clients.qbittorrent - INFO - qbittorrent.py:489 - Removed torrent from qBittorrent: 654d3e2df2edcda58fc6d998e256431bded28998

[alexhb1]

@mrskizzex Shelfmark is working as intended here, you shouldn't use the same directory for both your download client and the final shelfmark destination path. https://github.com/calibrain/shelfmark/blob/main/docs/configuration.md

[mrskizzex]

I misunderstood that part then, I will change download locations so they are different. But it doesn't change the fact that Shelfmark despite selecting file organization None. Keep original filenames from the source., Is pulling the files from the torrent's folder and putting them haphazardly to the download directory. I also tried using Rename and Organize. Create folders and rename using templates. Do not use with ingest folders. but it just resulted in: /Christopher Paolini/Eragon 1-10 from the above example.

I think there should be a clear message that "None" doesn't mean "None" when it comes to multipart downloads and also that true No-OP mode as @tldev suggested is needed.

[alexhb1]

That again sounds like it's working as intended. Shelfmark always moves / copies / hardlinks files from any source. The "None" organisation mode just does nothing to organise files in your chosen destination, it just takes whatever's downloaded and deposits them in your chosen directory and keeps original filenames. If this is multiple files, then it'll do the same.

I guess the reason it looks haphazard is because they're being deposited into the same folder the download client uses, which isn't recommended.

[mrskizzex]

If it was just preserving what was downloaded then there wouldn't be an issue, but you can't tell me that this is expected and desired outcome by anyone:

1. I search for Lord of the Rings in Shelfmark
2. I find a nice 13.3gb compilation of Tolkien with decently organized paths in the torrent folder:

3. I click download and after downloading it when it's selected:

• NONE it results in all files randomly in a download folder:

• Rename and Organize set to default {Author}/{Title}:

I don't understand the reasoning behind it. Sure this logic works when the torrent is for one book but It's not always the case.
No-OP mode would fix this.

For now I'll just jerry rig something up with custom scripts that will just move the original torrent file which remains untouched after shelfmark copies the files to it's download folder and afterwards i delete everything from that folder.

[alexhb1]

It works how it does because Shelfmark has a totally different model of files and processing than to what I think you're looking for, which does work great for most people.

Torrents and usenet are one equal part of a completely agnostic pipeline for outputting files. You have a destination folder that is separate from the original download location, the source reports back a download path, and the pipeline scans for valid book/audiobook files and delivers them with your custom processing options. The source directory layout isn't part of the approach because source directories aren't a thing for most sources. There's options to accommodate torrents via hardlink and custom directories/renaming, while also fully working for direct / IRC downloads.

This also looks like a multi-book download which isn't something Shelfmark handles - it's very purposefully a single book search -> download tool.

If you do want to move and rename with the original names in this way, use the {OriginalName} option in your path template. If you're looking for a truly different model, then that's what the custom script feature is available to help with.

[mrskizzex]

Thank you for explaining. I will try to do something with custom scripts except this bug i just found with them :) and I will deal with how Shelfmark handles audiobook downloads by just ignoring it and deleting whatever it copies to it's final download folder.

{OriginalName} wouldn't work because it would result in what selecting NONE does.

I understand that while I now know what Shelfmark was designed to do > Single book - single download, at least for audiobookbay downloads it's not really good approach because they often put up compliations of books.

Thanks again for clear explanations how Shelfmark works. I have nothing else to add and just end with a desire to add that option of doing absolutely nothing with the files after download except passing them to a custom script.