Usage

Smoke workflow defaults

The bundled workflow lives at samples/workflows/LTXLongAudio_CustomNodes_SmokeTest.json.

Current checked expectations:

Item	Value
Frame folder default	samples/input/frames_pool
Audio widget default	ltx-demo-tone.wav
App mode inputs	Frames Folder, Source Audio Upload, Segment Seconds, Random Seed
App mode output node	LTXVideoCombine

Sample assets

Tracked sample assets stay intentionally light:

• samples/input/frames_pool ships quarter-resolution stills at 688x384.
• samples/input/demo_frames ships tiny debug stills at 192x108.
• samples/input/ltx-demo-tone.wav is the tracked fallback audio for scripted smoke runs.

The tracked default stays ltx-demo-tone.wav. If you keep a longer local sample named HOWL AT THE HAIRPIN2.wav, the API smoke runner prefers it and stages it under the tracked upload name.

Input discovery behavior

The upload and folder nodes intentionally search both the ComfyUI input directory and the repository sample directory:

• audio dropdowns can expose files from the runtime input directory and samples/input
• image dropdowns do the same for still images
• folder dropdowns include both ComfyUI input folders and samples/input/* directories
• upload-oriented COMBO widgets keep a blank first option so the App mode surface can start cleanly

Publishing checks

Use the static checker before you publish a workflow update:

uv run python scripts/check_workflow_layout.py \
  samples/workflows/LTXLongAudio_CustomNodes_SmokeTest.json \
  --require-all-nodes-in-groups \
  --require-app-mode

Then use the live smoke script when a real ComfyUI install is available:

uv run python scripts/run_comfyui_api_smoke.py \
  --workflow samples/workflows/LTXLongAudio_CustomNodes_SmokeTest.json \
  --comfy-root /path/to/ComfyUI