Platform hub

Codex Skills

Find SKILL.md files and related workflows that appear tailored for Codex-style agent use and Codex-compatible setups.

Want an agent-native computer in the browser? Try HappyCapy.

Cloud sandbox for AI agents · No setup · Run autonomous workflows from your browser

Explore HappyCapy

Affiliate link — we may earn a commission at no extra cost to you.

Why browse by platform

Codex-oriented skills deserve their own platform hub because users increasingly want workflows that fit Codex-specific environments rather than a generic skill concept. Even when a SKILL.md file is transferable, platform hints still matter because they shape how install paths, task triggers, or sidecar scripts are expected to behave.

A Codex platform page helps surface those nuances while still keeping the directory useful for comparison. Users can browse by platform first, then narrow by category to find frontend, backend, debugging, or testing skills that are more likely to feel native in a Codex-centered workflow.

It is also a meaningful SEO angle. Searches for Codex skills, Codex agent workflows, or Codex SKILL.md patterns are more specific than generic skill directory queries, which makes them a strong fit for structured platform pages.

Benefits of this platform filter

  • Highlight skills that are more likely to fit Codex workflows without extra adaptation.
  • Create a clear browse path for users who know their preferred agent ecosystem already.
  • Improve platform-specific discovery without fragmenting the underlying category model.
  • Support ranking for narrower searches around Codex-compatible skills.

How to use codex skills

  1. Look for explicit Codex mentions, installation instructions, or repo conventions that suggest the skill was written with Codex in mind.
  2. Use the detail page preview to judge whether the workflow assumptions match your environment.
  3. Cross-check the platform with the category so you know both how and where the skill is likely to help.
  4. If a skill feels portable, keep it in a generic bucket, but prefer platform-specific ones for smoother adoption.