Vidmoat

Vidmoat is an agent-first video editor built around the idea that most short-form and talking-head content should be produced by describing the outcome instead of scrubbing a timeline. Users upload raw footage and the Moat AI agent transcribes the audio at word-level, then handles the mechanical parts of editing: cutting silence and filler, syncing karaoke-style captions, colour-grading, removing backgrounds, and reframing the same source into vertical formats for TikTok, Instagram Reels, YouTube Shorts, and horizontal YouTube in one pass. A text-to-edit workflow lets creators trim the video by deleting words from the transcript, which is often faster than working with clips for interview, podcast, or tutorial footage. The one-click shorts feature auto-detects highlight moments in a longer video and exports a batch of vertical clips ready for distribution. Beyond the in-app agent, Vidmoat exposes a Model Context Protocol (MCP) server so external assistants such as Claude, Cursor, or GitHub Copilot can drive the editor programmatically, which is unusual for consumer video software and opens the door to scripted pipelines where an LLM generates a plan (script, cuts, captions, thumbnails) and hands the render off to Vidmoat. Typical users are solo creators, podcasters, coaches, and small marketing teams who publish weekly and would rather spend the time on script and hook than on rearranging b-roll. It is not aimed at cinematic colourists or narrative editors who need frame-level control and node-based grading.