# Kinocut

A local-first video editing MCP server, Python library and CLI that lets AI agents trim, caption and repurpose video with FFmpeg, backed by typed tools and quality checks instead of invented FFmpeg flags.

Kinocut is an open-source video editing MCP server, Python client and kino CLI that gives AI agents a guardrailed local video-editing surface. It wraps FFmpeg, with optional Hyperframes and Whisper extras, behind typed tools, preflight validation and quality and release checkpoints, so agent-produced edits can be inspected before publishing.

It targets tasks such as trimming, captioning and repurposing footage into short-form video formats, and records a Video Receipt for each operation so results are traceable rather than based on invented FFmpeg flags. It works with Claude Code, Cursor and any other MCP client, and runs entirely on the local machine with FFmpeg required on the PATH.

It is written in Python, licensed Apache-2.0, and published to PyPI and the MCP registry.

Website: https://kinocut.dev/