image-to-cinematic-video

What problem does it solve?

This Skill solves the problem of turning a still image (or a prompt) into a coherent multi-scene cinematic video concept without manually coordinating prompts, scene continuity, frame extraction, stitching, and hosting.

Core Features & Use Cases

• End-to-end cinematic pipeline: generates/ingests a source image, produces multiple video scenes, extracts continuity frames, stitches clips into one final MP4, and records a reproducible manifest.
• CHAIN and PARALLEL continuity modes: supports sequential “one long take” continuity (via last-frame chaining) and montage-style parallel generation (shared first frame).
• Production-validated workflow: uses generate_image with authenticated byte fetching, uploads only to uguu.se for reliable serving, applies the verified Seedance filename handling workaround, and uses FFmpeg xfade for final transitions.
• Use cases: storyboard pre-visualization, marketing-style product reveals, cinematic opener concepts, and briefings for video AI tools like Runway/Kling/Sora via structured outputs.

Quick Start

Use the image-to-cinematic-video skill to create a 3–5 scene cinematic video by chaining scenes from the last frame continuity of the generated/uploaded starting image.