content / skill

OpenMontage

Turn your AI coding assistant into a video production workflow that researches, scripts, generates assets, edits and renders.

Repository owner calesthio

Not rated yet
0 comments

Like 0 Save

Published 11 Oct 2026Listing updated 12 Oct 2026

At a glance

Purpose: Turn your AI coding assistant into a video production workflow that researches, scripts, generates assets, edits and renders.
Runs in: Local agent with shell
Operating systems: macOS, Linux, Windows
Needs: Runtime: Python 3.10+, Node.js 18+ and FFmpeg; Needs internet: Yes; Needs accounts or API keys beyond your agent: Not stated
Files and data: Reads files: Yes; Writes files: Yes; Sends data to other services: Yes
Tool type: Skill
Delivery: From GitHub
Version: Latest on the main branch
Access: Code on GitHub
Compatibility: According to the README, it works with Claude Code, Cursor, Copilot, Windsurf and Codex, and with any AI coding assistant that can read files and run code.
External costs: The repository is licensed under AGPL-3.0, and the README states no price for the software itself. According to the README, every API key is optional, and a zero-key path uses free tools such as Piper TTS, open footage sources, Remotion and FFmpeg. The README names keys for services such as Atlas Cloud, Kling, Suno, ElevenLabs, OpenAI, Google, xAI, HeyGen and Runway, and for an image and video gateway. The README gives example costs per video, which depend on the providers you configure. It also describes cost estimates, spend caps and approval thresholds. Your coding assistant or model provider may also charge for usage.
Permissions: The software acts through your coding assistant with the access you give it. According to the README, the agent calls Python tools, runs FFmpeg and renderers, writes project files and final renders, and opens a local Backlot board. The agent runs live web searches during research, retrieves stock and archive footage, can analyze a reference video you give it (a YouTube video, Short, Reel, TikTok or local clip), and may call cloud APIs with the keys in your .env file. The README says it asks for your approval at creative decision points and includes spend caps and per-action approval thresholds.
Usage terms: GNU Affero General Public License v3.0, as stated in the repository. See its LICENSE file for the exact terms.

Code on GitHub

calesthio/OpenMontage

AGPL-3.0

Python67,579 stars8,624 forksLast pushed 12 Oct 2026 UTC

World's first open-source, agentic video production system. 12 production pipelines, 100+ tools, 700+ agent skill and production-knowledge files. Turn your AI coding assistant into a full video production studio.

Topics: agent agentic-ai ai claude copilot cursor elevenlabs ffmpeg

Public GitHub data fetched 13 Oct 2026 UTC · 0 days old

Skill folders in this repository (139)

Folders that contain a SKILL.md file, from GitHub's file list on 13 Oct 2026 UTC.

And 39 more in the repository.

View on GitHub ↗

What it does

OpenMontage is a video production system that runs through your AI coding assistant. According to the README, you describe a video in plain language and the agent handles research, scripting, asset generation, editing and final composition. The README says there is no code orchestrator: the assistant reads pipeline manifests and Markdown skill files and calls Python tools.

The README lists pipelines such as Animated Explainer, Animation, Avatar Spokesperson, Cinematic, Clip Factory, Documentary Montage, Hybrid, Localization and Dub, Podcast Repurpose, Screen Demo and Talking Head. Each follows the stages research, proposal, script, scene plan, assets, edit and compose, with human approval at creative decision points. It describes more than 100 tools and says it supports both free or local options and paid cloud providers.

The README also describes starting from a reference video, a local Backlot board that shows production progress, Remotion and HyperFrames for composition, and cost estimates before generation. The repository holds many skill folders for the media tools and libraries it uses, such as Remotion, FFmpeg, GSAP and Three.js.

Who it is for

People who use an AI coding assistant and want it to produce videos such as explainers, documentary montages, product demos or short clips.

Example tasks

Make a 60-second animated explainer about how neural networks learn.
Make a 75-second documentary montage about city life in the rain using real footage only.
Turn a reference YouTube Short into a video on a new topic, with cost estimates first.
Make a product launch teaser for a fictional smart water bottle called AquaPulse.
Run make demo to render zero-key demo videos.

Getting started

According to the README, clone the repository, change into the OpenMontage folder with cd OpenMontage and run make setup. Then open the project in your AI coding assistant and describe the video you want. Without make, the README gives a manual route that creates a Python virtual environment, installs requirements, runs npm install in the remotion-composer folder, installs piper-tts and copies .env.example to .env. API keys go in a .env file and are optional. The README has the exact commands.

Limitations

According to the README, you need Python 3.10 or later, FFmpeg, Node.js 18 or later and an AI coding assistant such as Claude Code, Cursor, Copilot, Windsurf or Codex. The README says every API key is optional and that more keys unlock more tools. Local GPU video generation needs make install-gpu and settings in the .env file. The README says the zero-key path uses Piper TTS narration, free or open footage sources, Remotion, HyperFrames and FFmpeg, and that the agent asks for your approval at creative decision points. The repository can change at any time; check its README before you install.

Data practices

According to the README, the free path uses Piper for offline text-to-speech, the Backlot board is local and builds from project files, and the agent saves checkpoints, decision logs and cost snapshots as JSON. The README names cloud APIs for images, video, speech and music that you enable with keys, and says the agent runs live web searches during research. What each service receives depends on the services you enable. What your coding assistant and model provider receive depends on them.