- Home
- Skills
- Agents & Orchestration
- Video Studio Skill
Video Studio Skill
Professional Remotion-based video production pipeline for vertical reels, vlogs, and automated motion design.
Free
Works with the AI Agents you already use
Video Studio Skill
Example session with this skill installed
I have three raw camera takes in assets/clips and a voiceover track in assets/audio. Build a vertical 2K reel with a lo-fi bed and calm captions. Use the night-lift script if the footage is dark.
- Read your context and instructions
- Compiled the video studio
I have initialized the project, scaffolded the Remotion environment, and processed your assets. The audio has been cleaned using DeepFilterNet, and a procedural night-preset lo-fi bed has been generated. The footage was analyzed for people zones to ensure correct 9:16 framing, and I applied the shadow lift to the two darker clips.
Reel Studio Processing Report
| Step | Status | Details |
|---|---|---|
| Ingest | Complete | 3 clips (04:12 total), 1 VO track |
| Voice | Restored | SNR: +18dB, DeepFilterNet applied |
| Crops | Verified | People tracked in shots 1-3, framing set to 9:16 |
| Timeline | Generated | 58s duration, 7-frame dissolves, punch zooms active |
| Captions | Ready | 18 pages generated, Russian (ru-RU), bleeps mapped |
| Loudness | Passed | -16.2 LUFS, True Peak -1.1 dBTP |
| QA Check | Clean | No digital silence or dropped frames detected |
Next steps
- Preview the layout in Studio:
cd reel && npx remotion studio. - Review
work/stills/qa1/to verify caption placement and face safe zones. - Run the final render:
scripts/render.sh . Reel2K out/project_v1_2k.mp4.
Connects securely to your tools. The creator never sees your data.
About this skill
Production-grade video editing is often a slow, manual process involving fragmented tools and inconsistent audio levels. Reel Studio automates the transformation of raw footage, screen recordings, or CapCut drafts into polished 2K vertical reels using Remotion. It handles the heavy lifting of audio restoration, transcript-based cutting, and motion design while strictly adhering to professional craft standards.
What it does
- Automated Montage creates finished videos from CapCut drafts, raw takes, or voiceover recordings.
- Audio Restoration cleans speech using DeepFilterNet, matches loudness to -16 LUFS, and manages multi-mic setups.
- Intelligent Framing uses Apple Vision to track subjects and maintain 9:16 crops from landscape sources.
- Motion Graphics generates calm phrase captions, motion scenes, and layout-aware overlays that never cover faces.
- Quality Assurance performs automated glitch scans, loudness checks, and dropped-frame detection before rendering.
How it works
- Initialize a project folder and scaffold the Remotion environment using absolute paths.
- Ingest material including camera clips, recordings, and LUTs for log footage.
- Process audio and video stems, including transcript generation and face zone analysis.
- Generate the timeline and motion scenes through Python scripts that drive Remotion components.
- Render the final 2K or 1080p output with a full QA report on audio peaks and frame integrity.
Frameworks & tools
This skill utilizes Remotion for programmatic video, Whisper for transcription, DeepFilterNet for audio cleaning, and BiRefNet for subject matting. It requires Python 3, Node.js, and FFmpeg, with specific optimizations for macOS (MLX) and Linux.
Why this beats prompting it yourself
Manual prompting lacks the precise control over frame-level timing and audio signal processing required for professional delivery. This skill implements a rigorous pipeline that prevents common errors like digital clipping, overlapping UI elements, or improper HDR-to-Rec.709 conversions.
Use cases
- Devlog Production turning screen recordings and voiceovers into structured technical reels.
- Event Vlogs multi-camera assembly with synchronized lavalier microphone audio.
- CapCut Migration moving mobile drafts to a high-resolution Remotion render pipeline.
- Series Generation batch processing a pool of material into a consistent set of episodes.
Known limitations
Windows support is restricted to WSL2 environments. Subject matting (BiRefNet) runs on CPU in Linux/WSL2, which is significantly slower than native macOS execution. Requires specific CLI tools for cover generation.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Agents & Orchestration.
Windows requires WSL2. DeepFilterNet and Whisper require specific Python environments. Face tracking and matting are optimized for Apple Silicon; other platforms see slower CPU-based execution.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 8 days ago
- Free with an Agensi account
- Accounts
GitHub
Hugging Face
Pytorch
Nodesource
Deno
Astral
- Software
Pythontsc
Node.jsremotioncreate-video
Git
Permissions
Allowed Hosts