Vox Style Collage Video

Agent skill for turning scripts, beat lists, or fact lists into Vox-style paper-collage explainer videos.

Demo

This skill is intended for Claude, OpenClaw, and Hermes Agent style skill runtimes. It is built around the Unsora MCP toolset from Unsora, which is required for image, video, music generation, and optional social posting.

Prerequisites

• A compatible agent runtime: Claude, OpenClaw, or Hermes Agent
• Unsora MCP connected and authenticated from tryunsora.com
• Available Unsora credits for image, video, and optional music generation
• ffmpeg and ffprobe available locally for assembly, timing, and mixing
• A voiceover source, either supplied audio or a separate session TTS tool

Unsora is deliberately not used for narration because Unsora has no TTS. The voiceover must come from a supplied audio file or another TTS tool connected to the session.

What It Does

The skill defines a narration-first production workflow using:

• Unsora MCP image generation for hero frames and reference sheets
• Unsora MCP video generation for limited-frame cutout animation
• Unsora MCP music generation for editorial underscoring
• External voiceover or session TTS for narration
• ffmpeg for timing, assembly, and final mix

Contents

• SKILL.md - main skill instructions and workflow
• references/style-dna.md - visual system, prompt templates, and style checks
• references/pipeline.md - generation, voiceover, ffmpeg, and delivery pipeline
• references/beat-mapping.md - narration segmentation and shot mapping

Usage

Install or copy this folder into your agent's skills directory, then invoke it for Vox-style, editorial collage, or paper-collage explainer video work.

Before using it, confirm that Unsora MCP exposes the required tools:

• create_image / wait_for_image
• create_video / wait_for_video
• create_music / wait_for_music
• Optional: get_subscription, get_credits, create_post, list_posts