what is generative-media-skills fr?

Generative media skills is a library of reference material that teaches AI coding assistants how to produce image, video, audio, voice, and music content professionally. Instead of being a piece of software you run, it is a collection of over 150 skill files organized into 25 categories, meant to be loaded by tools like Claude Code, Codex, GitHub Copilot, Cursor, and similar agent systems that support Agent Skills. Each skill focuses on one part of media production: choosing between different AI providers and models, applying craft like cinematography, lighting, sound design, and editing, building finished deliverables such as ads, product videos, podcasts, or game trailers, and running checks on the final output such as loudness levels, captions, or file checksums. Skills also cover careful concerns like rights, consent, provenance, and localization before something ships. Every skill package includes a SKILL.md file with instructions, decisions, and examples for the agent to follow, along with an EVAL.md file used internally to score how well the skill was followed. Some skills also include optional scripts, reference documents, or asset files, so a task can move from being described in text to running a script when precision matters. The project can also connect to a separate related project called OpenMontage for orchestrated media workflows, though generative-media-skills itself works as a standalone library. To use it, a person points their AI agent at the repository and tells it to load whichever skills the task calls for, then describes the deliverable in a normal prompt, such as making a short product launch video. The project is written mostly in Python and is released under the MIT license, meaning it can be used freely, including for commercial projects, as long as the copyright notice is kept. The README documents the full folder layout and lists an instructions file, AGENTS.md, that other agent-specific instruction files point back to.