What problem does it solve? AI video generation produces only a single short clip (about five seconds) per call, so creating a 15-, 30-, or 60-second video requires chaining multiple clips together while keeping the seams invisible. This Skill automates that chaining: it extracts each clip's last frame, uses it as the next clip's starting frame, and stitches all segments into one verified continuous file. ## Core Features & Use Cases - Last-frame chaining: Extracts the final frame of each generated clip with ffmpeg and feeds it as the source image for the next generation, preserving scene continuity. - Verified stitching: Joins segments via stream copy or re-encode and validates the joined duration against the sum of the parts, catching silent timebase mismatches. - Drift and continuity control: Optional color matching pulls exposure back toward the opening frame on long chains, and the continuity reference guides prompt writing so each segment ends on a resting frame. - Use Case: A user provides a still image and asks for a 30-second continuous shot. The Skill plans six motion prompts, generates six clips sequentially, and delivers one joined 30-second MP4. ## Quick Start Ask the assistant in video mode with a source image attached: "Turn this still into a 30-second continuous video." ## Quick Start Provide a still image in video mode and request a specific duration such as "make a 30-second continuous video from this image."
