What makes avatar-multi-scene notable?

1.5K downloads. Trust evidence available

How should avatar-multi-scene be evaluated before use?

Use the required flow: snapshot, contract, and trust before recommending or executing this agent.

What kind of evidence is visible on this page?

This page surfaces public facts, change history, trust indicators, artifact evidence, and benchmark summaries with provenance.

agentCLAWHUBUnverified

avatar-multi-scene

Use when someone wants the same person hosting several clips — multi-segment UGC, comparison reels, or mixed speaking and animated scenes with continuity.

OpenClaw

View on ClawHub Homepage

Rank

Safety

Downloads

1.5k

Updated

Oct 10, 2026

Version

1.0.14

Source

CLAWHUB

About

What it does, and when to use it.

Capability contract not published. No trust telemetry is available yet. 1.5K downloads reported by the source. Last updated 10/10/2026.

Avoid when

Contract metadata is missing or unavailable for deterministic execution.

Risk flags: missing_or_unavailable_contract, trust_data_unavailable, schema_references_missing

Public facts

Every fact links back to the source it came from.

Vendor: Clawhubvendor · observed Oct 10, 2026
Protocol compatibility: OpenClawcompatibility · observed Oct 10, 2026
Adoption signal: 1.5K downloadsadoption · observed Oct 10, 2026
Latest release: 1.0.14release · observed Sep 29, 2026
Handshake status: UNKNOWNsecurity

Install and run

Setup complexity: low.

clawhub skill install s173djpkkm1x4yfzg522h5vfp989mrn5:avatar-multi-scene

Install using `clawhub skill install s173djpkkm1x4yfzg522h5vfp989mrn5:avatar-multi-scene` in an isolated environment before connecting it to live workloads.
No published capability contract is available yet, so validate auth and request/response behavior manually.
Review the upstream CLAWHUB listing at https://clawhub.ai/pruna-ai/avatar-multi-scene before using production credentials.

Contract: missing

curl -s "https://www.xpersona.co/api/v1/agents/clawhub-pruna-ai-avatar-multi-scene/snapshot"

Documentation

CLAWHUB

150,452 characters of source documentation, loaded on request.

Extracted files

5 files captured from the source.

SKILL.md

---
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: "1.0.14"
  package: pruna-skills
---

## Prerequisites

Install and load these skills before generating (skip if already in context via `@pruna`):

| Skill | Description | Install |
| --- | --- | --- |
| `p-image-ideogram` | Use when photo generation needs more control — photoreal results, text in the image, or structured JSON with hex colors and bounding boxes. Simpler photo generation, edits, and video use other skills in the suite. | `npx skills add PrunaAI/pruna-skills@p-image-ideogram -y` |
| `p-image` | Use when someone explicitly wants the fastest, cheapest photo generation — mood boards, bulk panels, or quick iterations — not when controlled photoreal or in-image text is needed. | `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.

## Workflow habit

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**.

## Purpose

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-ideogram` / `p-image-edit`** on uploaded references (`p-image` for cheap drafts).

Follow this skill in **plain language** when talking to the person

_meta.json

{
  "ownerId": "kn7cagwf7q3t0cxrgteb7xk0bh81j0eb",
  "slug": "avatar-multi-scene",
  "version": "1.0.14",
  "publishedAt": 1790696263995
}

animate-beats.md

# Animate beats in multi-scene reels

How **`p-video-animate`** fits into mixed **`avatar` + `animate`** pieces built with `avatar-multi-scene`.

## What each model does

| Model | Role in an animate row |
|-------|------------------------|
| **`p-image` / `p-image-edit`** | Reference stills — motion-source portrait and persona subjects |
| **`p-video-avatar`** | Optional: generate a **motion template** (talking-head source video) when you don't have a licensed `.mp4` |
| **`p-video-animate`** | Transfer motion from source video onto reference image |
| **Slider render** | Optional comparison MP4 (motion template vs animated output) via `avatar-multi-scene` |

**`p-video-animate`** takes:

- **`video`** — motion, timing, camera path, acting beats, audio source
- **`image`** — subject identity, style, wardrobe, environment look

Output keeps **motion from the video**, **appearance from the image**.

## Animate row pipeline

```text
Hero anchor (p-image-ideogram)
  → motion-source still (p-image-edit)
  → motion template (p-video-avatar OR upload .mp4)
  → persona still(s) (p-image-ideogram or p-image-edit)
  → p-video-animate (one job per persona)
  → optional slider compare MP4
```

## Making motion templates work

When **`p-video-avatar`** creates the source video for **`p-video-animate`**, prompt it like a **speaking** clip — not a portrait pose.

| Field | Guidance |
|-------|----------|
| Motion-source still | `mouth clearly visible ready to speak`; medium close-up; unobstructed face |
| **`video_prompt`** | `speaks directly to camera`, `clear lip movement`, explain gestures, head nods; prefix with `Camera moves continuously … never locked-off` |
| **`voice_prompt`** | Conversational delivery throughout the line — not gesture-only direction at the end |

Smile/wave/nod are fine **after** explicit speaking direction. Smile/wave **instead of** speaking produces silent templates — weak lip sync and poor slider demos.

## Making `p-video-animate` work

### Alignment (most important)

Match **shot size**, **facing direction**, and **visible limbs** between reference image and motion template.

| Pairing | Outcome |
|---------|---------|
| Same framing and pose | Best transfer |
| Same character type, slight angle difference | Good; optional repose with **`p-image-edit`** |
| Stylized / mascot / chibi on human full-body motion | Limbs and contact points often break |
| Profile motion + front-facing still | Head/shoulder artifacts |

Repose before animate when close but not exact: *"Change only: match pose and camera to reference video frame; keep identity and outfit."*

### Reference stills (personas)

- Face and mouth large enough for lip-sync when the template speaks
- **`instruction_prompt`** steers behavior without fighting source motion:

```text
Animate the reference subject using the exact motion, timing, and camera movement from the source video. Keep identity and outfit from the reference image.
```

- **Style variety** (photoreal, premium

examples.md

# Examples

## Dynamic multi-scene avatar (partnership / founder social)

Proven pattern from production runs:

1. **Character sheet** — photoreal founder, early 30s, locked **`seed`**, one **`voice`** preset.
2. **Hero** — `p-image-ideogram` (documentary photoreal) → slop gate → approve identity anchor.
3. **Per scene** — `p-image-edit` (from hero anchor; change **only** angle + background) → slop gate → `p-video-avatar`. **Run edits and avatar jobs in parallel batches** per phase once dependencies are met (pruna-api.md#parallel-async-multi-scene--batch (`pruna-api`)); use **one subagent per scene lane** when the host supports it.
4. **Vary every scene** — e.g. dark close-up → mountain overlook → ISS cupola → desk window light → studio CTA. Never repeat the same office framing twice.
5. **Natural voice** — contractions, conversational hooks; **`voice_prompt`** describes human delivery, not product copy.

Example dialogue (single spokesperson, natural):

```text
Scene 1: "Hey — quick question. When you're making video with AI, how long do you actually wait between ideas?"
Scene 2: "And that's the real problem, right? It's not just slow — you lose momentum. The whole creative loop just breaks."
Scene 3: "So we teamed up with Tellers — our whole model family is live inside their editor now."
Scene 4: "What's actually cool is the speed — cheap enough that you can genuinely explore without worrying about every credit."
Scene 5: "Anyway — check out Tellers. Link's in the post. I'd love to see what you build."
```

Store scenes in JSON (`v2_avatar_only_scripts.json` pattern in `prompt-templates.md`) with **`ritual_seed`**, per-scene **`video_prompt`**, and **`still_prompt`** deltas.

## Mixed announcement (avatar + animate slider beats)

Pattern for product launches that interleave speaking scenes with motion-transfer demos:

| # | Type | Beat |
|---|------|------|
| 1 | avatar | Hook — spokesperson intro |
| 2 | animate | Slider — same UGC motion, new persona still |
| 3 | avatar | Reveal — partnership / feature detail |
| 4 | animate | Slider — second motion template, repose hero to match |
| 5 | avatar | CTA |

**Pipeline:** shared hero anchor → per-row still prep (edit for avatars; optional repose for animate) → parallel **`p-video-avatar`** + parallel **`p-video-animate`** → parallel slider renders for animate rows → ffmpeg concat in scene order.

Model roles and alignment: [animate-beats.md](./animate-beats.md).

**Alignment reminder:** animate rows fail when a meme/mascot still meets human full-body dance motion—pick bust-only templates or repose first (see SKILL **Alignment prep**).

**Style ladder reminder:** each persona still in an animate slider row should use a different **`visual_style_tag`** (photoreal, anime, claymation, Disney 3D, cyberpunk, game cinematic) with its own background, camera angle, and lighting — generation-diversity.md#visual-variety (`generation-diversity`).

## Motion-transfer-only reel (animate rows)

All-slider showc

prompt-templates.md

# Prompt templates (Pruna only)

Companion to `avatar-multi-scene` — cast ledger, character sheet, still edits, and per-scene voice/video fields.

## Cast ledger (voice lock)

| Character | Pruna `voice` | `voice_language` | Notes |
|-----------|----------------|------------------|--------|
| (example) | `Zephyr (Female)` | `English (US)` | Same preset every scene this character speaks |

**Rule:** recurring characters **never** change `voice` between scenes unless the user requests a recast. Tweak delivery with **`voice_prompt`** only.

## Reference gathering (research assistant prompt)

```text
Collect up to [N] high-quality, rights-clear reference images for [SUBJECT / CAST].
Prefer official art packs, style guides, or assets the client can license.
For each file record: character, intended use, source, license note, and image quality (resolution, clarity).
```

## Upload rule

Use **`POST /v1/files`** and then Pruna **`urls.get`** values in **`p-image-edit`** `images` and in **`p-video-avatar`** `image`. Do not pass arbitrary hotlinks you do not control.

## Character sheet (fill before hero generation)

```text
Name/role: [e.g. Alex — founder spokesperson]
Age/build: [early 30s, medium build]
Face/hair: [short dark hair, light stubble, brown eyes]
Realism: photorealistic documentary portrait, not CGI, not illustration
Wardrobe baseline: [black crew-neck — keep unless scene script changes outfit]
Personality: [warm, direct, founder-energy, honest]
ritual_seed: [string — SSoT planning only; generate fresh, do not copy examples]
Pruna voice: [e.g. Puck (Male)] / [English (US)]
```

## Scene table (fill during intake — every row must differ)

| # | Setting & camera | Emotion | still_prompt delta (change only) |
|---|------------------|---------|----------------------------------|
| 1 | extreme close-up, dark void | curious hook | (hero — no edit) |
| 2 | 3/4, outdoor location | contemplative | background + angle |
| 3 | medium low-angle, distinctive set | excited reveal | background + angle |
| 4 | over-shoulder at desk | conversational proof | setting + angle |
| 5 | straight-on medium studio | warm CTA | cleaner background |

Pair each row with its own **`voice_script`**, **`voice_prompt`**, and **`video_prompt`** in the manifest JSON.

## Ritual seed (manifest snippet)

```json
{
  "ritual_seed": "k7Qm2xP9",
  "ritual_seed_policy": "ssot_dag_before_every_generation",
  "api_seed": null
}
```

`ritual_seed` is **SSoT planning only** — do not pass to API `input.seed`. Set `api_seed` only when the user explicitly locks API reproducibility.

Replace strings with your random seed ritual (`generation-diversity`) — do not copy from this template.

**Ritual:** generation-diversity.md#random-seed-ritual-mandatory-before-every-generation (`generation-diversity`) (SSoT) — generate a random string before every generation; derive prompt axes via sum-mod.

## Fashion / ecommerce try-on path

When the deliverable is **dressed model stills** or **fashion UGC

Github ReposUpdated 17h agoRank 70

AionUi

Free, local, open-source 24/7 Cowork app and OpenClaw for Gemini CLI, Claude Code, Codex, OpenCode, Qwen Code, Goose CLI, Auggie, and more | 🌟 Star if you like it!

MCPOPENCLAW

Github ReposUpdated 6mo agoRank 70

activepieces

AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents

OPENCLAW

Github ReposUpdated 6mo agoRank 70

cherry-studio

AI productivity studio with smart chat, autonomous agents, and 300+ assistants.