excel_VedioPrompt

Video prompts have more moving parts than image prompts - you're not just describing a subject, you're describing a shot: who's in it, what style it's rendered in, what the light's doing, where it's happening, how the camera's framed, and what mood it's going for. excel_VedioPrompt is Apt_Preset's answer to writing all of that by hand every time: a template-driven prompt builder aimed specifically at video generation (the dropdown options lean heavily on Wan's official shot vocabulary).

Unlike the pack's other prompt-builder nodes, which mostly just concatenate whatever you pick, this one actually does template substitution - and it documents its own mechanism with a concrete example, which is worth quoting because it makes the behavior unambiguous: type a template like "{light}，女孩在{scene}玩耍" (roughly, "{light}, a girl playing at {scene}"), pick "白天" (daytime) for light and "草坪" (lawn) for scene, and the node fills those two placeholders in to produce a complete sentence. That's a meaningfully different mechanism from a plain dropdown-concatenator: the template field itself (33 choices, including Wan's own official shot templates alongside category headers like specific shot types, low-angle framing, and camera options) defines the sentence structure, and the other inputs are the values that drop into its placeholders.

The inputs feeding those placeholders: roles is free text for who's in the shot. style (204 options) sets the overall rendering look, starting from Wan's own official style as the first real entry. light (110) and scene (192) cover time-of-day/lighting and location, the same large scene library the pack reuses elsewhere. camera (156) is genuinely detailed here - options aren't just "wide shot" but full framing descriptions like low-angle mid-shot crowd emphasis, or low-angle wide-shot group scale, useful if you're trying to get specific about composition rather than leaving it to chance. atmosphere (42) sets mood - playful, tender, tense, and so on. solar_term (22) is weather/sky condition. object and target are two more free-text fields the template placeholders can reference (per the node's own doc, {object} and {target} are valid substitution keys alongside {roles}, {style}, {light}, {scene}, {camera}, and {atmosphere}). There's also an optional multiline custom field for when none of the templates fit and you'd rather write the structure yourself.

Like excel_Prompter, most of this dropdown content is in Chinese - this is a China-based author's pack, and the video-prompt library follows the same pattern. The required language toggle (default on) is presumably a translation switch, but expect the raw dropdown labels themselves to be in Chinese regardless of how you set it.

Output. A single STRING, pos - the assembled, template-filled prompt, ready for a text encoder or one of Apt_Preset's own sampler-stack nodes.

Installing it. Same as the rest of the pack - search "ComfyUI-Apt_Preset" in ComfyUI Manager, or:

cd ComfyUI/custom_nodes
git clone https://github.com/cardenluo/ComfyUI-Apt_Preset.git

Run install.bat on Windows for dependencies, then restart. This node is pure text assembly - no models, no GPU work - so if the pack imports cleanly, it works immediately.

Troubleshooting. If your placeholder isn't getting substituted, double-check the exact key name against the node's own list ({roles}, {style}, {light}, {scene}, {camera}, {atmosphere}, {solar_term}, {object}, {target}) - a typo'd brace or an unsupported key just gets left in the output as literal text rather than throwing an error. If the whole pack fails to load, that's the documented pack-wide IMPORT FAILED issue traced to a missing dependency elsewhere, not this node - check the ComfyUI console log at startup and reinstall via Manager's "install missing custom nodes."