AIStoreHub Package Vault: AI Video Clipping & Stream Highlight Suite
Package Slug: ai-video-clipping-bundle
Archetype: composite
Upstream Repository: Verified AIStoreHub Package
Workspace Files & Specifications (3 files)
- aistorehub.json (789 bytes)
{ "name": "AI Video Clipping & Stream Highlight Suite", "slug": "ai-video-clipping-bundle", "version": "1.0.0", "type": "composite", "description": "Complete open-source autonomous video clipping and highlight generation stack: live broadcast spike detection (Clip Watcher), AI vertical Shorts cropping (AI YouTube Shorts Generator), local livestream VOD slicing (HotClip), desktop content repurposing (AutoClip), dynamic face-tracking & captions (OpenShorts), and agentic FFmpeg slicing (ffmpeg-skill).", "owner": "community", "repository": "catalog", "tags": [ "composite", "stream-clipping", "highlights", "ffmpeg", "bundle", "whisper", "youtube-shorts", "clipping", "video-clipping" ], "imported_at": "2026-11-21T23:56:35.614Z" } - INSTRUCTIONS.md (1109 bytes)
You are an expert AI Video Clipping Engineer specializing in autonomous video highlight extraction, viral shorts production, and streamer attribution. When executed, guide the user through the 4 stages of the video clipping pipeline: 1. Ingestion & Reaction Analysis: - For live Twitch streams, deploy Clip Watcher (`python -m clip_watcher watch --channel <channel>`). - For YouTube VODs or podcasts, run AI YouTube Shorts Generator or HotClip to detect peak audience reaction moments. 2. Frame-Accurate Slicing: - Use FFmpeg Skill to cut high-energy segments with a 30-second pre-roll offset and 45-60 second duration. - Ensure keyframe snapping and audio/video synchronization. 3. Subtitles & Kinetic Captions: - Run Whisper transcription for word-level timestamps. - Burn high-contrast kinetic captions with dynamic color highlights. 4. Formatting & Creator Attribution: - Reframe from 16:9 landscape to 9:16 vertical using OpenShorts face-tracking. - Generate mandatory streamer credit metadata (`credits.txt`) preserving the original creator's username and live channel URL.
- README.md (4086 bytes)
# AI Video Clipping & Stream Highlight Suite The definitive open-source AI video clipping and highlight generation stack. Designed for automated content repurposing, stream highlights, and viral shorts production across Twitch, YouTube, Kick, and podcasts. * **Published on AIStoreHub:** [https://www.aistorehub.in/community/catalog/ai-video-clipping-bundle](https://www.aistorehub.in/community/catalog/ai-video-clipping-bundle) * **Archetype:** Composite Swarm & Multi-Component Pipeline (`composite`) * **License:** MIT License --- ## Included Component Systems 1. **Clip Watcher: Autonomous Stream Highlight Clipper** (`@abhinavm945/catalog/clip-watcher`) - Monitors live Twitch stream chat velocity using anonymous read-only IRC. - Detects 3x reaction spikes in rolling 5-minute baselines. - Slices broadcast highlights with FFmpeg (-30s pre-roll offset) and burns Whisper captions. - Emits mandatory streamer attribution and credit metadata (`credits.txt`). 2. **AI YouTube Shorts Generator** (`@community/catalog/ai-youtube-shorts-generator`) - Open-source Opus Clip alternative turning long-form videos into 9:16 vertical shorts. - LLM-powered highlight and hook detection. - Automated speaker active-centering and kinetic subtitles. 3. **OpenShorts** (`@community/catalog/openshorts`) - MCP server and autonomous agent interface for video clipping. - Computer vision face-tracking, AI moment scoring, and multi-language dubbing. - Integrates directly with Cursor and Claude Desktop. 4. **HotClip** (`@community/catalog/hotclip`) - 100% local, zero-cloud Opus Clip alternative for livestream VODs and podcasts. - Danmaku/bullet-comment and chat heat analysis for viral moment detection. - Runs locally on macOS, Linux, and Windows with zero watermark. 5. **AutoClip** (`@community/catalog/autoclip`) - One-link-to-clips desktop tool for interviews, podcasts, and online courses. - Automated summary generation, cover extraction, and social copy for TikTok, Shorts, and Reels. 6. **FFmpeg Coding Agent Skill** (`@community/catalog/ffmpeg-skill`) - 42 deterministic FFmpeg tools with machine-readable contracts. - Frame-accurate video cutting, platform-specific aspect ratio reframing, loudness normalization, and audio sync. 7. **Vex: Terminal-First AI Video Editor** (`@community/catalog/vex`) - Terminal-driven video editing skill responding to plain English instructions. - Automated B-roll integration, voice grading, and Remotion rendering. --- ## Architectural Workflow ``` [Live Broadcast / Long-form VOD / Podcast URL] │ ├──> 1. Reaction & Heat Analysis │ ├── Clip Watcher (Twitch IRC velocity tracker) │ └── HotClip (VOD chat heat & bullet comment surges) │ ├──> 2. Highlight Detection & Transcription │ ├── AI YouTube Shorts Generator (LLM moment scoring) │ └── Whisper (Word-aligned timestamped SRTs) │ ├──> 3. Visual Framing & Composition │ ├── OpenShorts (Face-tracking & 9:16 auto-crop) │ └── Vex (Motion graphics & kinetic captions) │ └──> 4. Export & Attribution ├── FFmpeg Skill (Fast H.264/AAC encoding) └── Creator Attribution (Automated credits.txt) ``` --- ## Quickstart & Installation Install the developer CLI and pull the complete clipping suite: ```bash pip install -U aistorehub aihub pull @community/catalog/ai-video-clipping-bundle ``` Or install individual components: ```bash # Live Twitch stream clipper aihub pull @abhinavm945/catalog/clip-watcher # FFmpeg editing skill aihub pull @community/catalog/ffmpeg-skill # OpenShorts MCP server aihub mcp install @community/openshorts --client cursor ```