AI-powered Auto-Clipper: automatically detects highlight segments from YouTube channels, Twitch/Kick live streams (via audience-retention data, chat spikes and timestamped comments), clips them with FFmpeg, generates subtitles using Whisper, and uploads them to a target YouTube channel. Includes a FastAPI + Celery backend, React + Tailwind frontend
git clone https://github.com/VadlapatiKarthik/autoclipper.gitThe Auto-Clipper skill is an advanced AI-powered tool designed to streamline the process of creating highlight segments from various video platforms, including YouTube channels and Twitch/Kick live streams. By leveraging audience-retention data, chat spikes, and timestamped comments, this skill automatically detects the most engaging moments and clips them using FFmpeg. Additionally, it generates subtitles with Whisper and uploads the final clips to a designated YouTube channel, making it an invaluable asset for content creators and marketers looking to enhance their video offerings. One of the key benefits of the Auto-Clipper skill is its ability to save significant time for users. Instead of manually sifting through hours of footage to find the best moments, users can rely on this automation to handle the heavy lifting in just around 30 minutes of implementation. This efficiency allows developers, product managers, and AI practitioners to focus on more strategic tasks, such as content strategy and audience engagement, rather than getting bogged down in the minutiae of video editing. This skill is particularly useful for content creators, social media managers, and marketing teams who regularly produce video content and need to optimize their workflow. By automating the clipping process, these professionals can quickly generate shareable highlights that resonate with their audience, driving engagement and increasing viewership. For instance, a Twitch streamer can use Auto-Clipper to create highlight reels from their live sessions, which can then be shared across social media platforms to attract new followers. The implementation of the Auto-Clipper skill is rated as intermediate, requiring a basic understanding of FastAPI, Celery, and React with Tailwind for the frontend. This skill fits seamlessly into AI-first workflows by automating repetitive tasks, allowing teams to leverage AI automation for enhanced productivity. As organizations continue to embrace AI technologies, the Auto-Clipper skill stands out as a practical solution for anyone looking to optimize their video content creation process.
[{"step":"Configure your Auto-Clipper settings","action":"Set up the FastAPI backend with your YouTube API credentials, Twitch/Kick channel credentials, and target YouTube channel details. Configure retention thresholds (e.g., clips must have >75% audience retention) and chat spike sensitivity (e.g., >20 messages per minute).","tip":"Use the React frontend to preview detected clips before processing. Adjust thresholds based on your channel's typical engagement patterns."},{"step":"Select the target stream","action":"Specify the live stream URL or ID from Twitch/Kick in the Auto-Clipper interface. The system will automatically pull audience retention data and chat logs.","tip":"For VODs (Video on Demand), use the 'Process Archive' feature to analyze past streams. This is useful for channels with irregular live schedules."},{"step":"Review and customize clips","action":"The AI will generate a list of potential clips with timestamps and descriptions. Use the frontend to preview each clip, edit descriptions, and merge/split segments as needed.","tip":"Prioritize clips with high retention peaks and chat engagement. Use Whisper's subtitle editor to correct any transcription errors."},{"step":"Process and export","action":"Click 'Generate Clips' to run FFmpeg commands for each segment. The system will create optimized .mp4 files with subtitles and prepare the YouTube upload package.","tip":"For large streams, enable Celery's distributed task queue to process clips in parallel. Monitor the backend dashboard for progress updates."},{"step":"Upload to YouTube","action":"Use the provided YouTube upload package (clips, subtitles, thumbnails) to schedule or publish the highlights. The package includes optimized titles, descriptions, and tags based on the clip content.","tip":"Schedule clips during peak audience hours (use YouTube Analytics to determine optimal times). Enable 'Chapter Markers' using the timestamps for better viewer engagement."}]
No install command available. Check the GitHub repository for manual installation instructions.
git clone https://github.com/VadlapatiKarthik/autoclipperCopy the install command above and run it in your terminal.
Launch Claude Code, Cursor, or your preferred AI coding agent.
Use the prompt template or examples below to test the skill.
Adapt the skill to your specific use case and workflow.
Generate a 30-90 second highlight reel from the latest [LIVE_STREAM_PLATFORM] stream of [CHANNEL_NAME] using audience retention data, chat spikes, and timestamped comments. Clip the segments with FFmpeg, generate subtitles using Whisper, and prepare the files for upload to YouTube. Include timestamps and a brief description of each clip. Target platform: [TARGET_PLATFORM].
Here are the auto-generated highlight clips from the latest 'TechGuruLive' Twitch stream (October 15, 2023) based on audience retention and chat spikes: **Clip 1: "The AI Breakthrough Everyone Missed"** - Timestamp: 00:12:45 - 00:15:30 - Duration: 2 minutes 45 seconds - Audience retention peak: 87% (highest in segment) - Chat spikes: 42 messages mentioning "AI" and "breakthrough" - Description: TechGuru demonstrates a real-time AI model generating code snippets from natural language prompts. The audience retention spikes when he reveals the model's accuracy rate (92%) and compares it to existing solutions. Includes viewer reactions in chat: "This changes everything!" and "Why isn't this getting more attention?" - Subtitles generated via Whisper (attached as .srt file) - FFmpeg command used: `ffmpeg -i input.mp4 -ss 00:12:45 -to 00:15:30 -c copy output_clip1.mp4` **Clip 2: "The Controversial Hot Take"** - Timestamp: 01:08:15 - 01:10:00 - Duration: 1 minute 45 seconds - Audience retention: 78% - Chat spikes: 67 messages with "hot take" and "controversial" - Description: TechGuru makes a bold claim that "most cloud providers will be obsolete in 5 years" due to edge computing advancements. The segment includes heated debate in chat with top messages: "Prove it!" and "This is either genius or insane." - Subtitles generated via Whisper (attached as .srt file) - FFmpeg command used: `ffmpeg -i input.mp4 -ss 01:08:15 -to 01:10:00 -c:v libx264 -crf 23 -preset fast -c:a aac -b:a 192k output_clip2.mp4` **Clip 3: "The Viral Fail Moment"** - Timestamp: 02:33:45 - 02:35:10 - Duration: 1 minute 25 seconds - Audience retention: 91% (highest overall in stream) - Chat spikes: 112 messages with "fail" and "epic" - Description: TechGuru attempts to live-debug a complex system but accidentally deletes the production database. The moment includes his shocked reaction, chat's "OH NO" flood, and his recovery process. Top chat messages: "This is the most entertaining debug session ever" and "I’ve seen this before… in a meme." - Subtitles generated via Whisper (attached as .srt file) - FFmpeg command used: `ffmpeg -i input.mp4 -ss 02:33:45 -to 02:35:10 -c:v libx264 -crf 18 -preset ultrafast -c:a copy output_clip3.mp4` **YouTube Upload Package:** - Three .mp4 files (1080p, H.264, AAC audio) - Corresponding .srt subtitle files - Thumbnail suggestions based on peak retention moments - Recommended title format: "[CHANNEL_NAME] | [Clip Title] | [Stream Date]" - Description template: "The best moments from [CHANNEL_NAME]'s [STREAM_TITLE] stream! Don’t forget to like and subscribe for more content like this. Timestamps: [LIST_TIMESTAMPS]" All clips are optimized for YouTube's algorithm with high retention segments and engaging hooks. The subtitles are timestamped to sync perfectly with the video content.
pre-order platform for b2b sales
Container platform for building, sharing, and running applications
Live streaming platform for gamers and creators
Customer data platform for collecting and routing events
Automate your browser workflows effortlessly
AI-powered voice transcription for sales teams
Take a free 3-minute scan and get personalized AI skill recommendations.
Take free scan