Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
Version History
ai-video-remix 0.1.2 - Converted this skill to an instruction-only/documentation format. - Added explicit notice that runtime source code must be cloned separately; no code is bundled. - Provided GitHub source and homepage links in metadata. - Updated Quick Start instructions to clarify cloning and setup process. - No code or functional changes included in this release.
What This Skill Does
Generates styled video remixes from a local ShotAI video library using natural language prompts. An LLM parses the request, selects a composition style, runs semantic search against indexed footage, and hands matched clips to Remotion for final rendering. Works without an LLM API key in heuristic fallback mode.
Replaces manual timeline editing by automatically searching, selecting, and arranging clips based on semantic understanding of both the request and the video content.
When to use it
- Cutting a travel vlog from vacation footage across multiple cities
- Assembling sports highlights from recorded game clips
- Creating a nature documentary-style montage from wildlife videos
- Generating a cyberpunk-themed edit from nighttime city footage
- Building an emotion-driven highlight reel from a personal video collection
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Make a travel vlog from my Tokyo trip footage
- 1Parses intent and selects TravelVlog composition
- 2Rewrites search queries to match Tokyo-specific content in the library
- 3Runs semantic search against ShotAI to find best shots per slot
- 4Extracts matched clips with ffmpeg and assigns visual effects via LLM
- 5Renders final MP4 with Remotion including location cards and captions
A styled travel vlog MP4 with location titles, transitions, and background music
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.