Video Pipeline Bundle

What This Skill Does

The Video Pipeline Bundle is a comprehensive automation toolkit designed for video content creators and developers. It serves as an end-to-end orchestration layer for video processing, moving files from raw footage to a polished, subtitled final product. By wrapping powerful command-line utilities like auto-editor, Faster Whisper, and FFmpeg into a single workflow, it allows users to perform complex video editing tasks without manual intervention. The skill handles automatic noise and silence removal, high-accuracy speech-to-text transcription with LLM-based error correction using MiniMax, and permanent subtitle burning.

Installation

Installation is streamlined via the OpenClaw environment. First, ensure you have system-level dependencies like ffmpeg and ffprobe installed on your machine. You can then install the skill using: clawhub install openclaw/skills/skills/leochens/video-pipeline-bundle. Once installed, use the built-in diagnostic tool by running python3 scripts/pipeline.py --check-deps to verify your environment. If dependencies are missing, run python3 scripts/pipeline.py --install-deps to automatically configure the required Python packages, including auto-editor and faster-whisper.

Use Cases

• Content Creation: Automate the editing of talking-head videos by removing "dead air" and silences automatically.
• Accessibility: Quickly generate accurate, hardcoded subtitles for educational or social media content.
• Bulk Processing: Batch process dozens of raw video clips for a video series, ensuring consistent styling and high-quality transcription.
• Archive Workflows: Streamline the conversion and tagging of raw video archives into manageable, searchable, and ready-to-publish files.

Example Prompts

1. "OpenClaw, use the video-pipeline-bundle to process the raw video in /home/user/footage.mp4, remove the silence, generate subtitles, and burn them into the final output."
2. "I need to transcribe the lecture video at /mnt/data/video.mp4 using the small model and verify the text accuracy with MiniMax. Please output the result to a new file."
3. "Run the full video optimization pipeline on all mp4 files inside the /projects/vlog_drafts folder and save the finished versions to /projects/vlog_ready."

Tips & Limitations

• Resource Usage: Video processing is computationally intensive. Whisper models vary in RAM requirements; for systems with limited memory, stick to the tiny model.
• API Keys: Ensure your MINIMAX_API_KEY is securely exported as an environment variable to prevent accidental exposure in your command history.
• FFmpeg Dependencies: This tool is a wrapper; ensure your FFmpeg installation is correctly mapped in your system PATH to avoid execution errors during the burn-in phase.