Debug Input

The most useful thing about this node is how little it does. Wire anything into it, run the graph, and it prints what it got to the console - that's the entire feature. No output to reconnect, no config to get wrong. When you're building a graph with custom nodes whose data shapes you don't fully understand, this is the fastest way to actually look at what's coming down a wire instead of guessing.

How it works

The input accepts anything - its type is the wildcard *, so ComfyUI won't stop you wiring in an IMAGE, a MODEL, a custom object from some other pack, whatever. On run, it prints the value to the console, and if what it received isn't a plain primitive (a number or string), it also prints a directory-style listing of the object's fields and methods. That second part is the real payoff: if a node you don't fully understand hands you back some custom Python object, Debug Input will show you what's actually on it, which is often faster than going and reading that node's source.

It's an output node with no outputs of its own - it's meant to sit at the end of a wire you've tapped off your real graph, not in the middle of one. Splitting a wire to also feed this node doesn't touch the rest of your pipeline.

The inputs and outputs that matter

• input (type *) - the only input, and it accepts literally anything you connect to it.

No outputs. This node terminates whatever it's plugged into; it doesn't pass data along.

How to install it

Via ComfyUI Manager: search "WAS_Extras", install, restart. Manually:

cd ComfyUI/custom_nodes
git clone https://github.com/WASasquatch/WAS_Extras

Restart ComfyUI. No dependencies - it's a print statement wearing a node.

Common issues & troubleshooting

Nothing shows up anywhere in the browser. That's expected - this prints to the console/terminal ComfyUI is running in, not to the web UI. If you're running ComfyUI locally, that's the terminal window you launched it from. If you're on a hosted service without a live console tail, this node's main value - reading raw Python output at runtime - won't be visible to you the same way; it's fundamentally a local-debugging tool.

You get a wall of text and can't find what you're looking for. For non-primitive inputs it dumps a full directory listing of attributes and methods, which can be long for a complex object. Skim for field names that look like actual data rather than dunder methods (__something__) - those are usually noise.

You want to debug mid-graph without breaking the flow to your next node. Just tap the wire: connect the same output to both Debug Input and whatever node was already consuming it. Since this node has no outputs, adding it never changes what downstream nodes receive.

The graph errors before you even see console output. If the node the wire is coming from is throwing, Debug Input never runs - the error is upstream of it, not in this node. Move your debug tap earlier in the graph to isolate where things actually break.