Smart File Transfer (Auto-Rename)

Most ComfyUI workflows end at the save node and expect you to shuffle the results around in a file manager afterward. Smart File Transfer is the "but what if the workflow did that itself" node: give it a source file, a destination, and it copies or moves the file with collision handling, directory creation, and a report of what it actually did.

The practical angle: you've got a pipeline that renders, and you want the output automatically organized into dated folders, or copied to a network drive, or moved out of the render directory into the "keep" folder - without you being the middleman. It's file management as a graph node, with sensible failure reporting instead of silent confusion.

The inputs that matter

The required ones are simple: source_path (the file), destination_path (a folder or full file path), operation (copy or move), create_directories (on by default - it'll make the destination folder tree for you), rename_pattern, and collision_handling. The optional custom_filename lets you rename on the way.

The two settings worth understanding up front:

• collision_handling - auto_rename (default) appends (1), (2) etc. to avoid clobbering; overwrite replaces; skip leaves the existing file alone. auto_rename is the safe default for "never lose a file," and it's why the node's display name says "(Auto-Rename)".
• rename_pattern - the style of the auto-rename suffix: parentheses (file(1).ext), underscore (file_1.ext), or dash (file-1.ext).

Under the hood it's stdlib shutil/pathlib - it validates the source exists and is a file, walks up the numbering until it finds a free name, and raises a clean error (returned as success: false) rather than a traceback when something's wrong.

Outputs

Three, all designed to be branchable:

• result_message - a STRING saying what happened (or why it failed)
• final_path - the actual path the file landed at, so you can wire it onward even when auto-rename changed the name
• success - a BOOLEAN you can gate on

That trio is the node's real value: in a fragile graph, you can check success and skip downstream work instead of blindly assuming the file moved.

What it isn't, and honest gotchas

It's a file transfer node, not a checksum-verified sync tool - the README blurb mentions checksums and "organized directory structures," but what the code actually does is copy/move with collision handling. Don't plan a backup pipeline on it.

Real gotchas: it's not a scheduler, so it runs when your workflow runs - if the source file doesn't exist yet at that moment, you get a clean failure, not a retry. And if your destination_path is a full file path rather than a folder, make sure it's what you meant; the node treats the destination as the file target when it looks like one. When in doubt, point it at a folder and let create_directories + auto-rename do the work.

Install

One of ~69 nodes in TrentNodes (TrentHunter82/TrentNodes):

# ComfyUI Manager: search "Trent Nodes"

# or:
cd ComfyUI/custom_nodes
git clone https://github.com/TrentHunter82/TrentNodes.git
cd TrentNodes && pip install -r requirements.txt

No extra dependencies beyond the pack's core. It's a small node, but the pattern - "let the graph finish the job instead of you" - is the kind of thing that quietly saves you twenty minutes a day once you remember it exists.