ControlNetApply_motorway_edition

This is the plain-vanilla ControlNet application node - the one from the original ControlNet-era graph: one conditioning, one loaded control net, one image, one strength. It bakes the control into the conditioning so the sampler obeys the structure you fed it. The _motorway_edition version is the same node, but both the inputs and the output live in the Motorway bus rather than on wires.

Why you'd use this over the Advanced variant

ControlNetApply is the simpler sibling. Where ControlNetApplyAdvanced takes positive and negative conditioning plus a VAE and handles both at once, this one takes a single conditioning and returns a single CONDITIONING. If your workflow doesn't use a negative prompt for control (or you're applying control only to the positive side), this is the lighter touch. It's also the node most historical ControlNet workflows were built around, so it's the familiar shape if you're porting one onto the Motorway.

The inputs that matter

• MOTORWAY 🚌💨 - the bus, required.
• INPUT_conditioning_key - key of the conditioning to apply control to, default conditioning.
• INPUT_control_net_key - key of the loaded CONTROL_NET, default control_net. Load it with ControlNetLoader_motorway_edition (or the plain core loader) and park it in the bus.
• INPUT_image_key - key of the condition image (canny edges, depth map, pose skeleton, whatever), default image.
• strength - the control weight, 0–10, default 1. The KB's standing advice: 0.3–0.7 for loose guidance, 0.8–1.2 for strict adherence, and for structure-heavy work let the condition drop off after composition forms.
• OUTPUT_CONDITIONING_key - where the conditioned result goes, default CONDITIONING.

Output is the Motorway. There's no start/end percent on this variant - that's what the Advanced version is for - so the control applies across the whole sample unless you range-limit it elsewhere with something like ConditioningSetTimestepRange_motorway_edition.

Installing

The whole pack, one shot:

cd ComfyUI/custom_nodes
git clone https://github.com/agilly1989/ComfyUI_agilly1989_motorway

Restart, or use ComfyUI Manager ("ComfyUI_agilly1989_motorway"). No pip deps; bring your own ControlNet model files and preprocessed condition image.

Caveats

Same pack-wide notes: active beta, "BIG BROKEN WITH ASYNC" README banner, and the _motorway_edition clones were commented out of __init__.py in the version I checked (uncomment ClonedNodeMapping if the node's missing). Keys are exact and case-sensitive - the hash_ error means a key isn't in the bus. And honestly, if your controlnet workflow uses a negative prompt (most do), the Advanced variant is usually the better Motorway fit since it handles both sides in one node. This one's for the simple, one-conditioning case.