openclaw-vision-storyboard

Inspects storyboard frames with bounded local vision inference to produce scene-indexed revisions.

1|Updated Sep 3, 2026

Search Tags:#ocr #video generation #storyboard #short-form video #frame analysis #vision inference #caption placement

Ready to run

npx skills add https://github.com/sabiscore/the-yap-engine --skill openclaw-vision-storyboard-sabiscore

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: openclaw-vision-storyboard
Source: https://github.com/sabiscore/the-yap-engine/tree/main/.agents/skills/openclaw-vision-storyboard
Command: npx skills add https://github.com/sabiscore/the-yap-engine --skill openclaw-vision-storyboard-sabiscore

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Storyboards and rendered short-form videos often contain composition flaws, misplaced captions, OCR errors, and continuity breaks that are hard to catch without visual inspection, especially on CPU-only machines where vision models compete for limited RAM. ## Core Features & Use Cases - Bounded Vision Inference: Runs exactly one admitted vision workload at a time with minimal frame sampling and resolution to fit constrained hosts. - Structured Visual Observations: Converts frame analysis into composition, subject/foreground/background, OCR, caption-safe region, continuity, and defect findings. - Storyboard Revision Feedback: Feeds observations back into the storyboard contract as concrete, scene-indexed revisions. - Use Case: Before rendering a TikTok short, sample key frames from each scene, detect that a caption overlaps the subject's face, and receive a deterministic revision moving the caption to a safe region. ## Quick Start Inspect the sampled frames from my storyboard and return scene-indexed observations with concrete revisions for caption placement and continuity.

Frequently Asked Questions about openclaw-vision-storyboard

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I check storyboard frames for caption placement issues?▼

Sample the minimum frames needed per scene, resize them to the smallest resolution that preserves evidence, and run vision inference to identify caption-safe regions. The output is scene-indexed observations with concrete revisions for caption placement.

How to run vision models on a CPU-only machine with limited RAM?▼

Run exactly one admitted vision workload at a time and never keep a vision model resident beside an active 7B model on constrained hosts. Release the vision model before starting another heavyweight model stage.

Can vision inference detect OCR errors in video frames?▼

Yes, frame inspection extracts OCR text as part of structured observations covering composition, continuity, and visual defects. When OCR confidence is low, the skill reports insufficient evidence rather than inventing text.

What are the limitations of frame sampling for video inspection?▼

Frame sampling trades coverage for resource safety, so issues between sampled frames may be missed. The workflow stops when each sampled scene has enough evidence for a deterministic revision or an explicit insufficient-evidence result.

When should I not use local vision inference for storyboard review?▼

Avoid it when frames cannot be sampled at a resolution preserving required evidence, or when a vision model cannot be loaded without displacing an active heavyweight model. In those cases, defer inspection until resources free up.