ScheduleToCond
ScheduleToCond — turning a parsed prompt schedule into conditioning, for graph builders
- clip
- prompt_schedule
- CONDITIONING
This one's a plumbing node for people composing graphs at a lower level than "type a prompt into a text box." Its required prompt_schedule input is typed PROMPT_SCHEDULE - a custom type specific to this pack, meaning it's not something you type in directly; it's the output of an earlier parsing step somewhere else in your graph (the same kind of parsed-schedule object PCLazyTextEncode builds internally when it turns your text into a chain of PCTextEncode + SetConditioningTimestepRange nodes). ScheduleToCond is the piece that takes an already-parsed schedule and actually renders it into standard CONDITIONING your sampler can use.
Why this is separate from PCLazyTextEncode at all
PCLazyTextEncode bundles the whole pipeline - parse the text, build the schedule, realize it as conditioning - into one convenient node. ScheduleToCond exposes just the last of those steps as its own node, for situations where you're building a schedule some other way (composing it programmatically, combining schedules from more than one source, or working with an intermediate PROMPT_SCHEDULE object that another node in a more advanced graph produced) and want to hand it off to conditioning generation without going through the all-in-one text-box path.
This is not a node a beginner working from a single prompt text field will typically wire by hand. It's there for people building custom graphs on top of Prompt Control's internals rather than using its headline nodes directly.
Inputs and outputs
Required: clip and prompt_schedule (type PROMPT_SCHEDULE). Output: CONDITIONING, ready for a sampler exactly like any other text-encode output.
Installing it
Ships with the rest of the pack - no separate step:
cd ComfyUI/custom_nodes
git clone https://github.com/asagi4/comfyui-prompt-control
Or ComfyUI Manager, search "ComfyUI Prompt Control". No models required.
Honest framing
This is a genuinely obscure node - a single-digit-impressions, internals-facing tool that most people using this pack will never touch directly, because PCLazyTextEncode already does everything it does plus the parsing step. If you've landed here because you're trying to figure out what produces or consumes a PROMPT_SCHEDULE object, the honest answer is that the pack's public documentation doesn't spell out that lower-level API in the material available here - you're at the edge of what's documented, and reading the pack's own Python source (or asking in its GitHub issues) is the more reliable next step than guessing at the exact contract from the node schema alone.
Who actually needs this
If you're a beginner who found this page because you're trying to schedule a prompt in ComfyUI, the short version is: you almost certainly want PCLazyTextEncode instead, and you can stop reading here. ScheduleToCond is only relevant once you're building something that generates or manipulates PROMPT_SCHEDULE objects independently of the standard text box - for example, a custom node of your own that assembles a schedule programmatically from data rather than a hand-typed string, or a more elaborate graph that needs to combine or transform a parsed schedule before turning it into conditioning. That's a small, specific audience: node developers extending this pack, not end users writing prompts.
The general lesson this node is a good example of: a well-designed graph-generation tool like this one tends to expose its internal pipeline as separate, composable pieces (parse text → build schedule → render conditioning) rather than only as one opaque all-in-one node. That's good engineering - it's what makes the pack extensible - but it also means a handful of the nodes it ships exist purely to be that extension surface, and this is one of them.
Inputs (2)
| Name | Type | Default | Description |
|---|---|---|---|
| clip | CLIP | — | |
| prompt_schedule | PROMPT_SCHEDULE | — |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| CONDITIONING | CONDITIONING | — |