LC Sampler Configure Pipe Out

Pipes are how you stop a workflow from becoming forty wires. One bundle carries your model, conditioning, seed, and all your sampler settings down a single line, and consumer nodes pull out what they need. LC Sampler Configure Pipe Out is the pull-out side of that bargain for sampling values: it takes an LC_PIPE in and expands the sampler settings back into individual sockets.

No widgets. You can't configure anything here, and that's the design - this is an unpacker, the mirror image of the pipe-packing sampler node. If the rest of your graph runs on pipes, this is how you get plain values out at the edge without dragging the whole bundle around.

How it works

You feed it one required input:

• pipe - an LC_PIPE from LC Sampler Configure (pipe), LC Pipe (in/edit), or any pack node that emits the shared LC_PIPE type.

In return you get the full sampler settings as sockets: total_steps, cfg_1, denoise, step_swap, cfg_2, sampler_name, scheduler, and detailer_steps. It also passes the pipe through untouched on the first output, so you can keep chaining further Get/Set nodes after it - pull the values off for one consumer and let the bundle keep traveling.

Where it fits

Think of the LC pipe family as three roles: a node that packs values in (LC Sampler Configure (pipe), the "(pipe)" variant), this node that unpacks them, and LC Pipe In/Edit/Out for the general bundle. This one is specifically for the sampler subset, which is handy when your downstream sampler wiring wants individual typed sockets (INT, FLOAT, the sampler enum) rather than a dict you'd have to re-split.

The pipe pattern is the same context-bundle idea the KB's plumbing doc describes: one wire instead of twelve, bought at the price of not being able to see what's inside. If a workflow "looks right but samples wrong," check what version of the settings actually made it into the bundle - pipes hide that.

Install

Part of the LC123 pack - Manager (search "LC123") or:

cd ComfyUI/custom_nodes
git clone https://github.com/lonecatone23/ComfyUI_LC123_nodes

Restart after. No extra dependencies (ComfyUI's torch/numpy only), MIT licensed. Cloud-hosted ComfyUI won't have LC123 until you request it, so shared pipe workflows need local or a pack request.

Honest take: if you're not already on the pipe bandwagon, this node does nothing for you - it only makes sense once your graph is carrying LC_PIPE around. But if you've committed to that pattern, this is the clean exit ramp.