7-e1even
Featured
learn-agent
A collection of notes on building coding agents, derived from the production development of the desktop agent Reina, with each mechanism simplified into a zero-dependency, single-file Node.js demo.
generative-media-skills
Generative Media Skills is a library of 150+ research-backed, executable skill packages that let AI agents plan, direct, validate, and deliver professional-grade image, video, audio, 3D, and interactive media productions.
Stars
104
7-day growth
No data
Forks
15
Open issues
0
License
MIT
Last updated
2026-07-14
AI repository intelligence
FR-AI / ANALYSISWhy it is worth attention
It uniquely bridges the gap between generative model APIs and real media production by combining provider intelligence, cinematography craft, deterministic guardrails, and built-in evaluation—all in a portable, agent-friendly format.
Who it is for
- Developers building AI agents for media creation
- Media production professionals seeking automation
- Content creators and marketers needing scripted media workflows
- AI researchers studying agent-based media generation
Use cases
- Automating the creation of short promotional videos with shot plans and QA
- Producing product launch clips, social shorts, or explainer videos from a brief
- Generating and validating podcast audio, avatar videos, or real-estate media
- Building multi-step media pipelines that require provider switching and quality checks
Strengths
- Provides current, verified facts from provider docs and model cards
- Includes real production craft (cinematography, lighting, editing) not just API calls
- Uses deterministic scripts for guardrails instead of improvised code blocks
- Every skill has an integrated evaluation (EVAL.md) for subject-specific scoring
Considerations
- Requires a compatible agent runtime or framework (e.g., Claude Code, Codex)
- Executable resources may depend on external tools or services not bundled
- The breadth of skills may overwhelm simple use cases without progressive loading
README quick start
Generative Media Skills
The whole media production suite for your AI Agent
150+ research-backed skills across 25 categories for creating, directing, validating, and delivering image, video, audio, voice, music, 3D, avatar, and interactive media. It combines researched operating knowledge with executable guardrails so your AI agents can move from a creative brief to a usable deliverable with provider context, production craft, safety boundaries, and technical checks.
Use it with Claude Code, Codex, GitHub Copilot, Cursor, OpenClaw/Hermes-style agents, or any runtime that can load Agent Skills. Connect it to OpenMontage when you want the same production intelligence inside an orchestrated media workflow.
It is operating knowledge and executable production support that helps an agent make evidence-based media decisions and prepare shippable work.
Start in One Prompt
Point your agent at the repository and name the production surfaces it should load:
Use generative-media-skills.
Load the necessary skills required for the below task.
Create a 15-second premium product launch clip. Include the shot plan, generation prompts, sound direction, rights/provenance notes, and final QA pass.
The agent gets researched judgment from the skills and can use bundled resources when the task benefits from deterministic execution.
From Brief to Delivery
One repository covers the production chain:
- Choose the right system - compare current providers, model families, APIs, gateways, local runtimes, costs, limits, and data paths.
- Direct the work - apply cinematography, lighting, performance, production design, visual continuity, sound, music, editing, and accessibility craft.
- Build real deliverables - produce ads, launch films, social shorts, explainers, game trailers, product imagery, podcasts, avatar videos, real-estate media, and more.
- Run deterministic checks - inspect media, captions, 3D assets, ComfyUI graphs, HyperFrames timelines, gateway plans, loudness, checksums, and QA reports.
- Ship with evidence - handle provenance, rights, consent, localization, platform requirements, manifests, and final delivery QA.
The agent gets judgment where judgment matters and executable validation where prose is not enough. When a skill includes executable
Description
Research-backed agent skills for premium image, video, audio, voice, and generative media production across AI coding assistants.
Related repositories
Similar projects matched by category, topics, and programming language.
0xwilliamortiz
Featured
openclaude-improved
OpenClaude is an open-source CLI coding agent that runs on any platform and supports a wide range of LLM providers, offering the same tools and workflows as Claude Code.
makecindy
Featured
cindy
Cindy is an open-source AI agent that runs locally on your machine, integrates multiple AI harnesses and models, and provides memory, skills, and automation to perform real work in your projects and apps.