mcp-video

> **Part of the StudioMeyer MCP Stack** — Built in Mallorca 🌴 · ⭐ if you use it

A note from us

We have been building tools and systems for ourselves for the past two years. The fact that this repo is small and has few stars is not because it is new. It is because we only just decided to share what we have built. It is not a fresh experiment, it is a long story with a recent commit.

We love building things and sharing them. We do not love social media tactics, growth hacks, or chasing stars and followers. So this repo is small. The code is real, it gets used, issues get answered. Judge for yourself.

If it helps you, sharing, testing, and feedback help us. If it could be better, an issue is more useful. If you build something with it, tell us at hello@studiomeyer.io. That genuinely makes our day.

From a small studio in Palma de Mallorca.

Features

Highlights

• 60fps frame-by-frame capture — Playwright screenshots every frame, ffmpeg encodes. Zero frame drops.
• Cinema easing curves — 16 easing options including `cinematic` and `showcase` for buttery smooth scrolling.
• Smart screenshots — Auto-detects 15+ UI elements (chat widgets, pricing sections, booking forms, etc.).
• Narrated videos — Provide a URL + script, get a professional video with synchronized AI voiceover.
• 22 LUT presets — Film-grade color grading (teal-orange, noir, vintage, cyberpunk, etc.).
• Social format export — One-click conversion to Instagram Reel, TikTok, YouTube Short, LinkedIn.
• Dual transport — Stdio (default) or HTTP mode for persistent microservice deployment.

Prerequisites

• Node.js >= 18
• ffmpeg and ffprobe (validated on startup, cross-platform)
• Playwright browsers (`npx playwright install chromium`)
• Optional: `ELEVENLABS_API_KEY` for ElevenLabs TTS
• Optional: `OPENAI_API_KEY` for Whisper captions and OpenAI TTS

If ffmpeg lives outside `PATH`, set `FFMPEG_PATH` and `FFPROBE_PATH` to the

absolute binary paths. Both env vars are honoured at startup AND at every

runtime spawn site.

Quick Start

With Claude Code (stdio)

{
  "mcpServers": {
    "video": {
      "command": "npx",
      "args": ["-y", "mcp-video"]
    }
  }
}

With npx

npx mcp-video

From source

git clone https://github.com/studiomeyer-io/mcp-video.git
cd mcp-video
npm install
npx playwright install chromium
npm run build
npm start

HTTP mode

# Start as HTTP microservice
npx mcp-video --http --port=9847

# Or via environment variables
MCP_HTTP=1 MCP_PORT=9847 npx mcp-video

Configuration

What You Can Build

Usage Examples

Record a website

Use video_record with type "cinema" to record https://example.com
with a smooth scroll and hover over the navbar.

Create a narrated explainer video

Use video_speech with type "narrated" to create a narrated video of
https://example.com with these segments:
1. "Welcome to our homepage" — pause on hero section
2. "Check out our features" — scroll to features
3. "Get started today" — hover over CTA button

Auto-caption a video

Use video_text with type "caption" to add auto-generated captions
to /path/to/video.mp4

Export for social media

Use video_compose with type "social-all" to convert
/path/to/video.mp4 to all social media formats.

Smart screenshot

Use video_screenshot with type "capture" to screenshot the chat widget
and pricing section on https://example.com

Architecture

src/
  server.ts            Entry point, 8 consolidated MCP tools
  lib/                 Logger, types, dual transport
  handlers/            Tool handlers (video, editing, post-production, tts, screenshots)
  schemas/             JSON Schema definitions for legacy tool format
  tools/
    engine/            Core engines
      capture.ts       Frame-by-frame recording (Playwright → PNG → ffmpeg)
      encoder.ts       ffmpeg encoding pipeline
      scenes.ts        Scene execution (scroll, hover, click, type, wait)
      cursor.ts        Visible cursor simulation
      smart-screenshot.ts  Element-aware screenshot engine
      tts.ts           ElevenLabs + OpenAI TTS with fallback
      narrated-video.ts    Full narration pipeline
      social-format.ts     Social media format conversion
      concat.ts        Video concatenation with transitions
      lut-presets.ts   22 cinema LUT presets
      ...and more

Development

npm run dev          # Start with tsx (hot reload)
npm run typecheck    # Type check
npm test             # Run tests
npm run check        # Verify ffmpeg/ffprobe installed

About StudioMeyer

StudioMeyer is an AI and design studio based in Palma de Mallorca, working with clients worldwide. We build custom websites and AI infrastructure for small and medium businesses. Production stack on Claude Agent SDK, MCP and n8n, with Sentry, Langfuse and LangGraph for observability and an in-house guard layer.

License

MIT

Credits

Built by StudioMeyer. Part of our open-source toolkit for AI-powered content creation.

• ai-shield — LLM security middleware for TypeScript
• agent-fleet — Multi-agent orchestration for Claude Code
• darwin-agents — Self-evolving agent framework