A curated, findable list of useful AI resources for working with Claude, Codex, and related tooling.
| Skill | Purpose |
|---|---|
frontend-design |
Better UI |
code-reviewer |
Catch bugs like a senior engineer |
document-skills |
Generate output as a file |
remotion |
Build videos in code |
skill-creator |
Build your own skill by describing what you want |
security-review |
Find security vulnerabilities in your code and how to fix them |
review |
Review your code |
claude-code-setup@claude-plugins-official |
improve coding capabilities |
/statusline create a persistant statusline showing my directory model and current context window usage |
status line to see context usage at all time |
| Repo | Description |
|---|---|
| public-apis/public-apis | Thousands of free public APIs |
| Leonxlnx/taste-skill | Better front-end design |
| fishaudio | Voice model |
| hardikpandya/stop-slop | A skill for removing AI tells from prose |
| microsoft/markitdown | Python utility for converting various files to Markdown |
| headroomlabs-ai/headroom | Reduce token usage without impacting performance |
| garrytan/gstack | 23 opinionated tools serving as CEO, Designer, Eng Manager, Release Manager, Doc Engineer, and QA |
| Anil-matcha/Open-Generative-AI | Self-hosted open-source AI image and video generation studio with 200+ models (Flux, Midjourney, Kling, Sora, Veo) |
| upstash/context7 | Get up-to-date code documentation loaded into LLMs |
| musistudio/claude-code-router | Locally running model gateway sitting between agents (Claude Code, Codex, ZCode) and upstream model services; unifies model/API key management, routing rules, log observation, and bot handoff |
| DietrichGebert/ponytail | A ruleset that makes your AI coding agent write the least code that works |
| webpro-nl/knip | Find unused files, dependencies, and exports in JavaScript/TypeScript projects |
| ferasbusiness666/ReconForge | Find security vulnerabilities in your app |
| Pantheon-Security/medusa | Security scanner with 40,000+ detection patterns for AI/ML, agents, and LLM applications |
| openai/codex-plugin-cc | Use Codex from Claude Code to review code or delegate tasks |
| nextlevelbuilder/ui-ux-pro-max-skill | Better UI/UX |
| bradautomates/claude-video | Gives Claude the ability to watch any video. /watch downloads, extracts frames, transcribes, and hands it all to Claude |
| nexu-io/open-design | Local-first, open-source Claude Design alternative. Native desktop app, 259+ skills, 142+ design systems; web/desktop/mobile prototypes, slides, images, videos; sandboxed preview; HTML/PDF/PPTX/MP4 export; works with 17+ CLIs |
| msitarzewski/agency-agents | A full AI agency of specialized expert agents, each with personality, processes, and deliverables |
| JuliusBrussee/caveman | why use many token when few token do trick — Claude Code skill that cuts 65% of tokens by talking like caveman |
| morphllm/morph-claude-code-plugin | Morph plugin for Claude Code - chagne reasoning token burn to code |
|
- claude.com/resources/use-cases
- Use skeleton loaders if items need time to load before they appear
- Leverage symlinks
claude agents— see all instances of Claude running locally in a single terminal- greta.sh — web app and website prompts
- vibeui.dev — UI prompts organized by component
- Type
/config> navigate tosessionRecap> set to enabled. Future sessions can then use/recap brew install trashand alias it torm. If AI deletes files, they go to Trash instead of being gone
I want the perfect prompt for [your goal].
Ask me questions one at a time until you have everything you need. Wait for my answer before the next question.
When you have enough, write the perfect prompt.
Then rate the prompt 1-10, tell me what's weak, and rewrite it stronger.
Generate five QA agents with 5 completely different personalities, archetypes, and personas based on who is likely to come to my site or use my app, and run a full QA test.