Indexes the coding-agent sessions already sitting on your disk — months of history from before you installed it — and serves them back over MCP. 3.5 GB searched in ~1.5 ms, no LLM, no embeddings.
Developer ToolsDatabases & Data
1 views
Orkas VideoStudio
by Orkas
MIT
videoffmpegtext-to-videosubtitlesmotion-graphicswhisper
Summary
Turns a coding agent into a video studio: describe a video in plain language and the agent writes an editable plan.json timeline, then renders, edits, generates and delivers the file.
Features
- Videos as an editable, diffable plan.json — change one line, re-render one segment
- Compose: script to HTML motion graphics to mp4, no paid keys
- Edit: cut, join, trim silence, de-filler, mix, burn-in subtitles, dub, localise
- Generate: talking-head and cinematic footage through your own provider keys (BYOK)
- Auto pipeline routes a cross-modal deliverable through one plan
- Deterministic delivery guard verifies the finished cut before it ships
- MCP tools mirror the `ovs` CLI one to one; SKILL.md packs for Claude Code and Codex
Installation
Set up this MCP server in your favorite AI agent — copy a ready-made configuration below.
Description
OrkasVideoStudio is deliberately not a black-box video agent. A video is expressed as a readable, diffable, re-renderable plan (plan.json) that both you and your agent can edit — change one line and only that segment re-renders. The agent supplies the judgement; the project supplies the knowledge of what makes a good video, the deterministic capabilities, and that editable intermediate representation.
You say something like "make a 60-second vertical explainer on vector databases with a Chinese voiceover and captions", and the agent reads the material, writes the timeline and produces the file.
Four production lines
- Compose — script to designed HTML motion graphics to mp4: explainers, kinetic typography, lower-thirds, data visualisation, title cards, transitions. No paid keys.
- Edit — cut, join, trim silence, remove filler words, mix, burn in subtitles, dub and localise footage you supply, plus highlight selection over long recordings. No paid keys.
- Generate — talking-head or cinematic AI footage and imagery through your own provider keys (OpenAI, Gemini, Doubao). No managed backend and no lock-in.
- Auto — when a deliverable spans more than one axis,
stage-planbuilds the edit decision list andstage-assemblewalks it, delegating each segment back to compose, generate or edit. A deterministic delivery guard (ovs plan promise-check) verifies the finished cut kept its promise — real motion rather than a silent slideshow — before anything ships.
The plan is the contract: "the footage in the middle, generate an opener, compose the stats, one voiceover over all of it" becomes a single editable file in which every segment, caption and narration line re-renders independently.
Running it
Needs Node 22 or newer and ffmpeg/ffprobe on your PATH. The renderer is a thin wrapper over HeyGen's hyperframes (pinned), ffmpeg and whisper.cpp. npm i -g @orkas/video-studio provides the ovs command; ovs doctor checks the prerequisites.
Agents pick it up three ways: as native SKILL.md knowledge packs in Claude Code and Codex, as MCP tools that mirror the CLI one to one, or through the self-describing CLI for any agent that can run a shell.
It began as the built-in video agent inside Orkas, the AI-team desktop app, and is now MIT-licensed and standalone.
Related MCP Servers
Microsoft's official MCP server for Azure DevOps - work items, repos, pipelines, wikis, test plans and iterations, reachable from any MCP client in plain language.
Developer Tools
Semantic code toolkit for coding agents — symbol-level search, editing and refactoring backed by language servers or the JetBrains platform.
Developer Tools
Official Hugging Face MCP server: search models, datasets, Spaces and papers on the Hub, read repository files, and call Gradio apps as agent tools.
Developer ToolsWeb & Search
1 views