Andrea Spoldi and Claude Sonnet 4.6
bdc3d577ad
Fix widgets_values positional alignment in workflow examples
...
ComfyUI reads widgets_values positionally including linked widgets.
Missing placeholder values for linked inputs caused remaining widget
values to shift — e.g. CompositionInferenceNode showed 0.4 in
ollama_url and 0.8 in model.
Add placeholders for linked widget positions in: CompositionInferenceNode,
MoodJsonUnpacker, PromptEnricher, CLIPTextEncode, EmptyLatentImage.
Fix applies to both example_comprehensive.json and example_composition.json.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com >
2026-06-10 19:20:59 +02:00
Andrea Spoldi and Claude Sonnet 4.6
16c58caad0
Add comprehensive example workflow using all node types
...
Full end-to-end pipeline: LoadAudio → AudioMoodAnalyzer + ClapAudioAnalyzer
→ CompositionInferenceNode (width/height) + MoodJsonUnpacker + PromptEnricher
→ CLIPTextEncode → KSampler → VAEDecode → PreviewImage.
22 nodes, 26 links, 3 groups (Audio Analysis, Image Generation Pipeline,
Previews). All 9 custom node types wired together. Model loaders use
placeholder names — swap to your own checkpoints.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com >
2026-06-10 18:52:57 +02:00
Andrea Spoldi
ae350f0f33
docs: add sample workflow, README docs and CLAUDE.md checklist for MoodJsonUnpacker + PromptEnricher
2026-05-20 18:00:40 +02:00
Andrea Spoldi and Claude Sonnet 4.6
66efa262ae
feat(composition): add example_composition.json workflow
...
Shows AudioMoodAnalyzer → CompositionInferenceNode pipeline.
width/height INT outputs wired directly into EmptyLatentImage.
composition_prompt previewed for wiring to CLIPTextEncode.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com >
2026-05-19 22:55:53 +02:00
Andrea Spoldi and Claude Sonnet 4.6
4dd1cad148
feat(clap): add example_clap.json sample workflow
...
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com >
2026-05-17 00:36:15 +02:00
Andrea Spoldi
2528f51843
feat: add example_animatediff.json workflow and README section for AnimateDiff formatter
2026-05-16 22:20:08 +02:00
Andrea Spoldi and Claude Sonnet 4.6
bdc28db530
fix: remove FL_Audio_Crop from timeline workflow — LoadAudio connects directly to Timeline node
...
The timeline node segments the full song internally; cropping defeats its purpose.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com >
2026-05-16 21:40:37 +02:00
Andrea Spoldi
858222e462
feat: add example_timeline.json workflow and README section for Timeline node
2026-05-16 21:29:21 +02:00
Andrea Spoldi and Claude Sonnet 4.6
30745dfd2d
workflow: replace text concatenation with dual CLIPTextEncode + ConditioningAverage
...
- Remove Text Concatenate, old positive CLIPTextEncode, and final PreviewAny
- environment_prompt → CLIPTextEncode #35 (conditioning_from)
- subject_prompt → CLIPTextEncode #36 (conditioning_to)
- ConditioningAverage #37 at strength 0.5 → KSampler positive
- summary PreviewAny kept for display only, no longer fed into conditioning
- Display PreviewAny nodes for mood_json, subject_json, environment_prompt,
subject_prompt, summary all preserved
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com >
2026-05-16 19:19:02 +02:00
