Agent skill · vm0-ai
ppt-avatar-video
Convert a PPT, PPTX, or PDF directly into a narrated video by preserving each page as a static slide, with an optional transparent talking avatar in a fixed side layout or adaptive overlay. Use for fast deck-to-video requests with voice only or avatar plus voice; do not use for cinematic intros, slide redesign, or tutorial screen recordings.
What it needs
About 8k tokens when loaded.
What this skill does
PPT Narrated Video Produce the final video quickly. This is a conversion workflow, not a motion-design workflow. Scope boundary Preserve every source page as a static image in the original order. Voice-only and overlay modes keep the slide full-frame; the two side-by-side modes fit it inside their fixed slide rectangle without cropping. Always include the selected narration. Overlay a transparent talking avatar only when the user selected one. Use hard cuts between pages by default. Add only a short dissolve when the user explicitly requests it. Do not select a design preset, construct an arc, search the HyperFrames catalog, choose blueprints or rules, redesign slides, or author per-slide animation. Do not invoke a general intro-video or motion-graphics workflow unless the user explicitly asks for custom motion. Default to direct final delivery. Pause for preview approval only when the user asks to review first. Preserve the requested configuration Use the supplied source, aspect ratio, and voice exactly. Treat avatar configuration as optional: never add or choose an avatar when the user supplied only a voice. When an avatar is supplied, preserve its ID, presenter placement, and presenter scale exactly. Do not list or search avatars or voices when compatible identifiers are already supplied. For avatar mode, map 16:9 to landscape, 9:16 to portrait, and 1:1 to square. Transparent JoggAI output requires --screen-style 3 --no-caption. If scale is absent, use a visible presenter width of 14% of the frame. Scale refers to the non-transparent avatar bounds, not the full video canvas. If an exact supplied avatar or voice is rejected, report that blocker. Never silently substitute another identity or voice. Select presenter placement The product exposes exactly three placement choices. Normalize their prompt labels to these manifest values: left — presenter on the left, slide on the right. …
How to use it
Reference it in AdaL, Claude Code, Cursor or any coding agent — nothing to install:
@skills vm0-ai/ppt-avatar-video