Menu
Use when someone wants the same person hosting several clips — multi-segment UGC, comparison reels, or mixed speaking and animated scenes with continuity.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Use when someone needs spoken narration or voiceover — explainer tracks, documentary lines, or voice to pair with generated video.
Use when someone wants an original AI song with vocals — sung lyrics, a style prompt track, or source audio for a music video.
Use when someone wants light instrumental background music — an ambient bed under dialogue or underscore for reels and explainers.
Use when someone needs word-level timestamps from audio — lyric alignment, cut-safe line boundaries, or caption source timing before burn-in with video-editing.
Use when crafting TTS, music, or bed prompts for any generative audio model — director style, song structure, and post-production layering.
Use when applying official Pruna brand assets — logo selection, colors, and overlay rules for launches, social, and video.
| name | avatar-multi-scene |
| description | Use when someone wants the same person hosting several clips — multi-segment UGC, comparison reels, or mixed speaking and animated scenes with continuity. |
| license | MIT |
| metadata | {"version":"1.0.8","package":"pruna-skills"} |
Install and load these skills before generating (skip if already in context via @pruna):
| Skill | Description | Install |
|---|---|---|
p-image | Use when someone wants a fast AI image — product shots, hero visuals, mood boards, or draft photos from a text prompt. | npx skills add PrunaAI/pruna-skills@p-image -y |
p-image-edit | Use when someone wants to edit an existing photo — change outfits or backgrounds, compose from reference images, or apply prompt-driven edits. | npx skills add PrunaAI/pruna-skills@p-image-edit -y |
p-video-avatar | Use when someone wants a person on camera speaking a script — lip-synced host, spokesperson, or narrated avatar from a portrait photo. | npx skills add PrunaAI/pruna-skills@p-video-avatar -y |
p-video-animate | Use when someone wants a photo to move like another video — motion transfer, dance remixes, or performance variations from a template clip. | npx skills add PrunaAI/pruna-skills@p-video-animate -y |
Or install the full suite once: npx skills add PrunaAI/pruna-skills@pruna -y
Follow each skill's Before generating / craft sections — do not restate guide content here.
In every reply, name `avatar-multi-scene` in backticks. Restate the user's continuity / UGC / host / segment goal in one line. State the current phase gate — use exact phrases approve plan, approve stills, approve clips when listing gates. Do not same-turn plan + paid video. Skip-review / burn-credits → follow generation-diversity Red flags.
Produce a coherent multi-scene piece stitched later with ffmpeg (Pruna does not ship a concat endpoint). Each beat is one of:
| Beat type | Model | Deliverable |
|---|---|---|
avatar | p-video-avatar | Talking-head clip from approved still + voice_script |
animate | p-video-animate + slider render | Motion-transfer clip, usually wrapped in a side-by-side or wipe comparison MP4 (motion template vs animated subject) |
Mix types in one announcement reel—e.g. avatar hook → animate slider demo → avatar CTA.
Visual continuity comes from Pruna p-image / p-image-edit on uploaded references.
Follow this skill in plain language when talking to the person requesting the video. Use natural, speakable copy in every voice_script.
Staged generation: generation-diversity · generation-diversity
| Resource | Path |
|---|---|
| Photoreal dynamic personas | image-prompting |
| Cast ledger, character sheet, voice/video prompts | prompt-templates.md |
| Animate rows, sliders, alignment | animate-beats.md |
| Examples | examples.md |
| Feedback discipline | generation-diversity |
| Slider (agent) | ffmpeg hstack / wipe — see Slider comparison below |
| Batch template | templates/batch.template.json |
| Phase | What to show | Proceed when |
|---|---|---|
| 0 — Plan | Scene table, read-through, cast ledger | approve plan |
| A — Stills | Hero + per-scene plates | approve stills |
| B — Video | Avatar / animate clips + sliders | approve clips |
| C — Assembly | Concat reel + optional bed | User accepts |
Open intake → generation-diversity clarification intake.
Do not call POST /v1/predictions until the user has answered and you have recorded the answers (use defaults only if the user explicitly opts in):
| Topic | Questions |
|---|---|
| Goal | What is the piece for (pitch, tutorial, trailer, episode)? Primary audience? |
| Scope | How many speaking scenes or beats? Approximate total runtime after assembly? |
| Cast | Who speaks, in what order? One character throughout or multiple? |
| Look | Aspect for stills and feel (9:16 / 16:9)? Avatar output 720p or 1080p? |
| Media source | Generate hero plates with p-image / edits vs upload-only references; user-owned motion templates for animate beats? |
| Voice | For each named character, pick one Pruna voice and voice_language and reuse it in every scene that character speaks. Any words that must be pronounced exactly (names, acronyms)? |
| Style | Agreed style bible line for all image prompts? |
| Character sheet | Per speaker: age range, wardrobe baseline, hair, skin/realism level, personality adjectives—record before hero generation (see Character sheet below). |
| Scene variety | Each scene must differ in camera angle, background/setting, and/or energy—no two consecutive scenes with the same framing and location unless the user asks. Plan visual_style_tag, setting_tag, camera_tag, lighting_tag per row — generation-diversity. |
| Ritual seed (SSoT) | Ritual seed at hero — generate and state a ritual string; log as ritual_seed; derive prompt axes via sum-mod. Do not pass ritual string to API seed. |
| References | Which files to upload; rights cleared? |
| Beat mix | Which scenes are avatar vs animate? All avatar, all animate, or mixed announcement? |
| Narrated B-roll cutaways | Optional p-video beats using scene-anchor triple (video-prompting) alongside avatar rows |
| Motion templates (animate beats) | Source .mp4 per animate row—owned/licensed? Match pose/framing to reference still? |
| Slider delivery (animate beats) | Comparison MP4 only, animated-only strip, or both? Canvas default 1920×1080. |
| Assembly | How clips will be joined and leveled (ffmpeg plan)? |
If anything material is unknown, ask before the first upload or prediction.
Maintain a cast table in the manifest: one Pruna voice + voice_language per recurring character — never swap presets mid-story unless the user requests a recast.
Before hero generation, fill a character sheet per speaker (age, face, realism, wardrobe baseline, personality, ritual_seed for planning). Templates and manifest JSON: prompt-templates.md.
Rule: New locations and styles = p-image-edit off the approved hero URL — not unrelated fresh p-image identity pulls.
Every piece needs a scene table — each row avatar or animate. Example columns and manifest JSON: prompt-templates.md · animate-beats.md.
P-Video-Animate animates a reference image using motion, timing, and camera movement from a source video. The better the subject's features, pose, framing, and proportions align with the motion template, the better the result.
| Alignment | Typical outcome |
|---|---|
| Same shot type, similar pose, similar scale | Clean motion transfer; slider demo reads instantly |
| Same character type, slightly different angle | Good with optional p-image-edit repose toward a template keyframe |
| Meme / cartoon / mascot on human full-body motion | Limbs, gait, and contact points may warp or slide |
| Tiny head / extreme proportions on dance or arm-heavy motion | Hands, legs, and depth cues often break |
| Reference facing camera, source subject in profile | Shoulder/head turn and occlusion artifacts |
Rule: Treat severe pose or proportion mismatch as a pre-flight risk. Repose with p-image-edit or pick a closer motion template before burning p-video-animate credits.
Alignment prep (per animate row):
p-image-edit from the hero anchor: "Change only: match pose and camera to reference video frame; keep identity and outfit."video-prompting on the pair before animate.Anti-patterns (all types): two identical office avatar scenes back-to-back; corporate brochure voice_script; human dance template + chibi meme still without repose; serial API jobs when scenes are independent; motion templates that prompt smile/wave only (avatar stays silent — see below).
When p-video-avatar generates a motion template (source video for p-video-animate), treat it as a speaking beat — not a portrait pose.
| Field | Requirement |
|---|---|
Motion-source still_edit | mouth clearly visible ready to speak — not passive smile only |
video_prompt | speaks directly to camera, clear lip movement, explain gestures, head nods — before any wave/smile close |
voice_prompt | Delivery throughout the line — not “wave energy at the end” only |
| Camera | Prefix: Camera moves continuously for the full clip — … never locked-off |
Silent motion templates break slider demos and animate transfers. Prompt templates: prompt-templates.md. Full animate pipeline: animate-beats.md.
| Pattern | Structure |
|---|---|
| Interleaved | avatar hook → animate demo → avatar proof → animate demo → avatar CTA |
| Slider-heavy | N animate slider rows → final avatar CTA on hero |
End product launches with a speakable avatar CTA unless the user opts out.
Complete the ritual seed step in generation-diversity before hero prompt work. Log ritual_seed in manifest; reuse only on same-brief slop retry.
Character continuity = approved hero plate URL + cast descriptor — not API seed. Pass api_seed in input only when the user explicitly locks reproducibility.
voice_script = speakable dialogue (contractions, short breaths). voice_prompt = performance direction only — never marketing copy or script text.
Good/bad pairs: prompt-templates.md.
For each recurring character:
p-image or upload. Run the slop gate on the hero before sign-off. Treat the approved file URL as the identity anchor.p-image-edit from that same source URL, plus the shared style bible and a short delta (“change only: …”).After intake is complete:
voice_script for avatar rows; motion templates + reference stills + alignment risks for animate rows; cast ledger; hero URL(s); chosen resolution; legal/CTA lines verbatim if supplied.POST /v1/predictions until the user explicitly confirms.Once the user confirms:
p-image → parallel per-scene p-image-edit → slop gates → approve stills.p-video-avatar (avatar rows) and p-video-animate (animate rows) via curl batches (pruna-api).Prefer one parallel lane per independent scene after the hero exists. Parent owns confirmation, manifest merge, and assembly.
p-video-avatar input.image — use an approved still URL from /v1/files that passed generation-diversity checklists.Hero: p-image (or upload) → slop gate → approve anchor
Scene N: p-image-edit(anchor) → slop gate → p-video-avatar
| Step | Model | Skill |
|---|---|---|
| Upload binaries | POST /v1/files | pruna-api |
| Style-locked stills | p-image, p-image-edit | p-image, p-image-edit |
| Talking clips | p-video-avatar | p-video-avatar |
| Motion transfer | p-video-animate | p-video-animate |
| Slider comparison | ffmpeg (below) | local |
Use PRUNA_API_KEY and the apikey header on every call. Async + parallel by default: batch all avatar jobs once approved stills pass slop; batch all animate jobs once motion + still URLs are ready; poll all get_url together.
| Phase | Parallel? |
|---|---|
Hero p-image → gate | Sequential |
Per-scene p-image-edit | Yes — all scenes |
| Slop gate | Yes — review in parallel |
p-video-avatar | Yes — all avatar rows |
p-video-animate | Yes — all animate rows |
| Slider render | Yes — all animate rows |
| Assembly | Sequential order only |
Rule: Never dispatch generation before user confirmation.
Side-by-side (motion template | animated):
ffmpeg -y -i motion_template.mp4 -i animated.mp4 \
-filter_complex "[0:v]scale=960:1080[l];[1:v]scale=960:1080[r];[l][r]hstack=inputs=2[v]" \
-map "[v]" -map 1:a? -c:v libx264 -c:a aac -shortest scene_compare.mp4
Or present both clips and let the user’s editor build a wipe slider. Paths can follow templates/batch.template.json (source, render, samples[]).
| Step | Action |
|---|---|
| 1–3 | Intake → speakable script → confirmation gate (no API until approve) |
| 4–5 | Upload refs → p-image hero per character → slop gate |
| 6–7 | Parallel p-image-edit scene stills → slop gate each |
| 8 | Parallel p-video-avatar (cast ledger voices, unique video_prompt per scene) |
| 9 | Parallel p-video-animate + ffmpeg sliders — animate-beats.md |
| 10 | ffmpeg concat ± optional bed — stable-audio-2.5 |
| 11 | Manifest: paths, prediction ids, slop notes, cast snapshot |
ffmpeg -y -f concat -safe 0 -i clips.txt -c copy reel.mp4
Shared ffmpeg recipes (concat, bed mix, sliders, export): video-editing.
Related skills:
| Skill | Description | Install |
|---|---|---|
avatar-single-scene | Use when someone wants one polished host-on-camera beat — a speaking person with intake and approval gates before generation. | npx skills add PrunaAI/pruna-skills@avatar-single-scene -y |
image-to-video | Use when someone wants one short film beat from images — a narrated scene, story moment, or cinematic B-roll with optional voiceover. | npx skills add PrunaAI/pruna-skills@image-to-video -y |
narrated-multi-scene | Use when someone wants a multi-part story with voiceover — episodic B-roll, chaptered promo, or several linked video scenes without on-camera dialogue. | npx skills add PrunaAI/pruna-skills@narrated-multi-scene -y |
p-video-animate | Use when someone wants a photo to move like another video — motion transfer, dance remixes, or performance variations from a template clip. | npx skills add PrunaAI/pruna-skills@p-video-animate -y |
video-editing | Use when assembling or polishing already-rendered clips with ffmpeg — concat, crossfades, burned captions and subtitles, text/logo overlays, before/after sliders, background music beds, platform export — or when composing a multi-layer HTML combination video with Hyperframes. Not for AI video generation, prompt craft, or model-based video edits. | npx skills add PrunaAI/pruna-skills@video-editing -y |