A curated collection of six agent Skills for Claude Code, Cursor, Codex and similar AI coding agents, covering web design, video, images, articles and local knowledge retrieval.
14.5K+443 this week1.7KJavaScriptMITUpdated 13 days ago
MengTo/Skills
A curated library of portable agent skill folders for designers and builders using Codex, Claude, Cursor and other AI coding agents.
- Stars
- 7,278
- Forks
- 848
- Open issues
- 10
- Last commit
- 2 days ago
What is it?
This repository is a collection of agent skills: folders containing a SKILL.md playbook, plus optional references, articles, scripts and demos. Each skill packages a repeatable procedure, such as turning a screen recording into a detailed prompt, capturing full-page screenshots, building Three.js game systems, or generating SVG illustrations, so an AI coding agent can load it and follow the same steps each time.
The library covers web design (landing pages, GSAP and scroll motion, WebGL effects, Tailwind CSS), browser game development and combat effects, 3D rendering, illustration styles, media sourcing, and workflow habits such as shipping changes with screenshots and tests. Skills are written in plain Markdown and are meant to be portable across Codex, Claude Code and Cursor rather than tied to one project.
The format is intentionally simple: no runtime or package to install, just files an agent reads before acting. Skills are organised by category under agent-skills/, and each includes triggers, default values and acceptance checks where relevant.
Key features
- 01Procedural SKILL.md playbooks that state when to use them and what to do first
- 02Portable folder format that works with Codex, Claude Code, Cursor and other agents
- 03Web design skills for landing pages, GSAP motion, WebGL and Tailwind CSS
- 04Browser game skills for Three.js systems, combat animation and playtesting
- 05Full-page capture and video-to-prompt workflows for reproducing reference designs
Use cases
- Reproducing a landing page or animation from a screen recording using a detailed prompt
- Giving a coding agent consistent rules for Three.js game or WebGL effects
- Standardising how an AI agent ships changes with screenshots and tests
- Building a shared prompt and workflow library for a design or development team
Quick start
- 1
Clone or open the repository and browse the agent-skills folder
- 2
List every skill file from the repo root
find agent-skills -name SKILL.md | sort - 3
Load the narrowest matching SKILL.md into your agent before acting
In collections
Topics
FAQ
- What is Skills?
- A curated library of portable agent skill folders for designers and builders using Codex, Claude, Cursor and other AI coding agents. This repository is a collection of agent skills: folders containing a SKILL.md playbook, plus optional references, articles, scripts and demos. Each skill packages a repeatable procedure, such as turning a screen recording into a detailed prompt, capturing full-page screenshots, building Three.js game systems, or generating SVG illustrations, so an AI coding agent can load it and follow the same steps each time.
- Is Skills free to use? What license does it have?
- Skills is open source under the MIT license. Check the license terms before using it in commercial products.
- Is Skills still maintained?
- Actively maintained. The last commit to Skills was on Nov 5, 2026.
- What can I use Skills for?
- Reproducing a landing page or animation from a screen recording using a detailed prompt Giving a coding agent consistent rules for Three.js game or WebGL effects Standardising how an AI agent ships changes with screenshots and tests Building a shared prompt and workflow library for a design or development team
- How do I install Skills?
- Clone or open the repository and browse the agent-skills folder List every skill file from the repo root: find agent-skills -name SKILL.md | sort Load the narrowest matching SKILL.md into your agent before acting
- Where can I get Skills?
- The source code and installation instructions are on GitHub: https://github.com/MengTo/Skills
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
A paste-in rules block that stops coding agents like Claude Code and Codex from over-defending, over-building and adding unneeded checks.
796+55 this week28PythonMITWell maintainedUpdated yesterday
A set of 24 open-source agent skills that steer Claude Code, Codex, Cursor and opencode toward less generic, more deliberately designed UI output.
217+8 this week13JavaScriptMITUpdated 3 months ago
A cross-tool skill that makes AI coding agents write less code and less prose, cutting output tokens and LLM API costs.
348+1 this week23JavaScriptMITWell maintainedUpdated 20 days ago
Agent Skills that make coding agents design distinctive frontends, using art direction, real photos and a browser self-check instead of generic AI landing pages.
111.2K+4.9K this week7.5KJavaScriptMITWell maintainedUpdated 3 days ago
Installs a spec-driven development workflow as Agent Skills for AI coding agents like Claude Code, Codex, and Cursor, covering requirements, design, tasks, and autonomous implementation.
3.8K+7 this week286TypeScriptMITUpdated 17 days ago