open-ribbi/velocut

AI-native, local-first video editor by Ribbi — runs entirely in the browser. Rust/WASM engine, WebGPU compositing, WebCodecs export; humans and LLM agents edit through the same JSON command protocol.

What it solves

Velocut provides a unified browser-based environment for video editing and 3D scene direction, allowing users to combine manual editing with AI-driven creation. It eliminates the need to switch between separate tools for 3D layout, video assembly, and AI video generation, integrating them into a single project runtime with a shared undo history.

How it works

The system consists of a multitrack video editor and a 3D Director. It uses WebGPU and WebCodecs for browser-side rendering and media processing. The project can be controlled manually via the UI or programmatically through an MCP (Model Context Protocol) server, which allows AI agents (like Codex) to create scenes, arrange objects, and control cameras. It also supports connecting to external AI video, speech, and music providers via a declarative configuration system, enabling users to generate video clips directly into specific timeline ranges.

Who it’s for

It is designed for creators and developers who want to use AI to accelerate video production and 3D scene composition, as well as those building AI-powered media tools using the provided SDKs.

Highlights

  • Integrated 3D Director: Create and animate 3D primitives, import GLB models, and pose characters within the video project.
  • AI-Agent Integration: Full MCP server support allows AI agents to perform complex scene creation and camera directing tasks.
  • Direct Timeline Generation: Generate AI video clips from text or first-frame prompts and place them directly into designated timeline slots.
  • Detailed History: A branching history system that attributes edits to specific users or AI agents, allowing for precise undos and revision checks.
  • Browser-Native: Media storage and rendering happen locally in the browser using IndexedDB and OPFS.

Related

  • Project
    cartesiancs/cartcut

    CartCut is a cross‑platform video editor that adds AI features (Whisper auto‑caption, Claude‑driven editing via an MCP bridge) to a layer‑based, keyframe‑animated workflow. It uses FFmpeg for rendering, supports 4K/8K, offers plugins, and is open‑source under MIT.

  • Project
    MartinDelophy/ai-video-editor

    A local-first, browser-based AI video editor featuring a multi-track timeline and WebGPU-powered AI for music, voiceovers, captions, and video restoration.

  • Project
    0xsline/OpenChatCut

    An open-source, local-first AI video editor that allows conversational agents to perform editable, multitrack timeline edits rather than just generating immutable videos.

  • Project
    NomaDamas/CozyClay

    CozyClay is a browser‑based pre‑visualisation studio built on Three.js/React‑Three‑Fiber. It lets you block a 3‑D scene, pose characters, animate cameras on a timeline, and export a keyframe pack that feeds directly into AI video generators (Seedance, Kling, Veo, MiniMax‑H3, etc.). Optional integration with the Kimodo motion‑generation service adds AI‑driven character animation. The app includes an AI chat panel (supporting ChatGPT, Anthropic, Gemini, OpenRouter, CLIProxy), an MCP server for external control, and a node‑based workflow canvas. Install with a single `npx cozyclay` command; no compilation needed.

  • Dispatch
    FableCut: A Zero-Dependency Browser Video Editor for AI Agents

    FableCut is a browser-based non-linear video editor that exposes its entire timeline as a JSON document, allowing AI agents to perform edits via MCP, REST, or direct file modification with live UI updates.