WeftCut

Product Introduction

1. Definition: WeftCut is a free, open-source, cross-platform desktop video editing application designed with native integration for AI agents via the Model Context Protocol (MCP). It functions as a professional-grade non-linear editor (NLE) that can be controlled programmatically.
2. Core Value Proposition: It exists to bridge the gap between creative intent and manual video editing labor by enabling AI agents to perform complex, granular editing tasks directly on a real timeline. Its value proposition centers on agent-native video editing, MCP integration for AI, and maintaining human-in-the-loop control over the automated editing process.

Main Features

1. MCP (Model Context Protocol) Integration: WeftCut exposes its entire editing interface as a set of tools over the standard MCP, a protocol for connecting AI applications to external resources. This allows any MCP-compatible client (like Claude Desktop or Cursor) to connect locally and execute commands. The integration includes a bundled Skill that informs the agent of WeftCut's full capabilities, enabling it to trim, split, add transitions, keyframe effects, insert captions, place markers, and undo actions.
2. Motifs - Code-Based Motion Graphics: This feature allows users or their AI agents to describe visual elements (titles, lower thirds, countdowns) which are then turned into "Motifs"—reusable, code-defined motion graphics templates. These Motifs render instantly on the timeline, enabling rapid iteration and consistent styling across projects by turning descriptive language into executable visual code.
3. Professional Desktop Editor Core: Underlying the AI features is a full-featured, local desktop video editor. Key technical capabilities include a multi-track timeline, Bézier curve editing for keyframes, per-clip effect chains, live chroma keying with color picker, support for subtitle files (SRT, VTT, ASS), global search (including Pinyin), and export to professional codecs including H.264, HEVC, 10-bit AV1, ProRes, and DNxHR.
4. Local Media Analysis: An optional, local analysis engine provides AI agents with contextual insight before editing. This includes shot detection, silence detection, speech-to-text transcription, and scene description, all processed on the user's machine without uploading media, giving the agent data to make informed editing decisions.

Problems Solved

1. Pain Point: The high cognitive load and repetitive manual labor involved in video editing, such as cutting silences, labeling scenes, or applying consistent text styles, which disrupts creative flow and momentum.
2. Target Audience: Content Creators & Video Professionals seeking efficiency; Developers & Technologists interested in programmable media; Teams using AI Agents (e.g., with Claude, Cursor) who want to extend their capabilities into visual media editing without switching contexts.
3. Use Cases: Automated Editing Workflows: An agent cutting silences from a podcast interview and tightening pauses based on transcription. Rapid Prototyping: Using natural language to instruct an agent to create and apply a complex title sequence via Motifs. Assisted Editing: A human editor focusing on creative pacing while delegating technical tasks like splitting clips at every shot change and applying effect keyframes to an AI assistant.

Unique Advantages

1. Differentiation: Unlike cloud-based AI video tools or traditional desktop editors, WeftCut uniquely combines a local, professional NLE with standardized agent control via MCP. It differs from automated online editors by offering granular, reversible control over a real timeline and keeping all data local. It surpasses scriptable editors by being built from the ground up for agent interaction.
2. Key Innovation: The agent-native architecture is its core innovation. Instead of bolting AI features onto an existing editor, WeftCut is designed so every UI action is inherently available to an MCP client. The checkpoint system for agent edits ensures all changes are visible, reviewable, and undoable, preserving ultimate human editorial control while automating execution.

Frequently Asked Questions (FAQ)

1. Is WeftCut a cloud-based AI video editor? No, WeftCut is a local desktop video editor that connects to your existing AI agent (which may use cloud models). The editing, rendering, and optional media analysis all occur on your computer; no video is uploaded to external servers for processing.
2. What AI models or agents work with WeftCut? WeftCut works with any AI application that supports the Model Context Protocol (MCP), such as Claude Desktop, Cursor, Windsurf, or custom clients. The app itself does not bundle or run any AI models; it provides the tools for your chosen agent to control.
3. Can I use WeftCut as a traditional video editor without an AI agent? Yes. WeftCut is a fully-featured, open-source video editor for Windows, macOS, and Linux. All its professional editing, effects, and export capabilities are available for manual use entirely independent of any AI agent functionality.
4. How does the "Motifs" feature relate to code? Motifs allow visual designs (like animated titles) to be defined and manipulated using code. This enables both users and AI agents to programmatically generate and modify motion graphics using a descriptive or script-based approach, making complex animations reusable and easily adjustable.
5. Is WeftCut really free for commercial use? Yes. WeftCut is released under the MIT license, making it free to use, modify, and distribute for personal or commercial projects. It bundles FFmpeg, which is licensed under LGPL/GPL for its components.