pyang5166
Community@pyang5166
pyang5166 provides editorial halftone paper-collage B-roll generation and cover image prompt design for article and video content creators.
Agent Skills by pyang5166
Showing 2 vetted skills indexed across 2 GitHub repositories.
Frequently Asked Questions About pyang5166
FAQPage SchemaWhat tasks can I accomplish with pyang5166's skills?▼
You can generate image prompts for article and video covers, and transform short voiceover scripts, opinion sentences, or abstract concepts into high-end editorial halftone paper-collage B-roll animations with a mandatory three-stage approval flow: visual metaphor, color still frame, then assembled animation.
Who should use the gbro-collage-broll skill?▼
Content creators and video editors who need collage-style visual metaphors for spoken scripts or abstract ideas. It triggers on requests like 'collage b-roll', '纸拼贴 b-roll', or turning a piece of copy into a collage animation.
What are the installation and runtime requirements for gbro-collage-broll?▼
It runs in a Codex environment with built-in image_gen for Gate 2. The bundled scripts/generate_video.py requires Python >= 3.10, google-genai >= 2.10.0, a configured GEMINI_API_KEY, plus ffmpeg and ffprobe. Complete the first-use environment self-check before running.
Which video model does gbro-collage-broll use by default?▼
The default video model is fixed to gemini-omni-flash-preview; Veo is no longer the default. The model only switches when the user explicitly specifies an alternative.
How does the approval workflow in gbro-collage-broll work?▼
It enforces three stages: first only the visual metaphor is proposed; after user confirmation, a color collage still frame is generated; after a second confirmation, Gemini Omni Flash generates first/last frames and assembles the final animation.