anthropics/claude-code-playground ? reverse-engineered prompt

Build me a GitHub repo that serves as a playground of Claude Code setups, skills, and demos, with everything organized into separate folders so people can browse examples easily.

I want it to feel like a curated collection, not a polished product. Include sections for Claude Code hooks, CLAUDE.md examples, settings examples, workflows, mods, reusable skills, launch demos, and showcase demos. Each entry should live in its own folder with a README that explains what it does, how it was built, how to run it, and any limitations. Add a reusable template folder and a simple checklist so new entries follow the same format.

Make the top level README explain what this repo is for, how to use it, and where to go if someone wants official docs or starter apps instead. Keep the repository easy to browse and self contained, with no setup needed at the root. If anything depends on current Claude Code docs or best practices, look up the latest info online first.