What it does
Tourwright writes a script, films a real Chromium browser with an animated cursor, highlights and camera zoom, and adds narration from text-to-speech voices. It outputs an MP4 with captions, chapters and a thumbnail, plus SRT and VTT subtitle files and a JSON manifest. Personal data can be masked, and sentences are cached so re-rendering does not re-synthesize them.
It suits teams that need to re-render a product tour after a UI change, or produce the same tour for many customers or languages from one script with variables. A check step in CI reports broken steps. AI agents can create videos through the MCP server or a Claude Code plugin. The tool can be used from the CLI, an HTTP API, the MCP server or a Node library.
Running it requires Node.js 22 or later, or Docker for a full service with the API, workers and Redis. Without Docker, the install step places ffmpeg, Piper voices, edge-tts and Chromium in a single folder without root or admin rights. The setup can scale from a single process on a Raspberry Pi to multiple workers on several servers.
Limitations and requirements
Requires Node.js 22 or later. The Redis queue and Claude-based narration are optional and need extra packages installed. Python 3.10 to 3.13 is needed for edge-tts, Pocket and Whisper; the Microsoft Store Python stub is not used. Snap and Flatpak browsers are skipped. The sharpest capture needs Xvfb on Linux; otherwise recording runs headless.
Safety
Our team has not reviewed this resource. Only automated rules checked it; resources with high-risk signals are not published. These are signals, not a guarantee that installing it is safe. Try it in a sandbox project first.
- Runs code or changes an assistant's behaviour
Sources
- Official MCP Registry first seen 13 Nov 2026 · last seen 13 Nov 2026 · its page on the source
Tourwright
Turn a script into a narrated product tour video: real browser, captions, voice, MP4.
v2.6.2As shown on registry.modelcontextprotocol.io when we last read it