Codex Skills — Testing skill for Claude Code
Five Codex CLI skills that give OpenAI Codex a Claude Code-style workflow (feature dev, PR review, browser testing, data engineering) plus an orchestrator that routes each task to official Codex plugi.
How to install Codex Skills
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open Ark310/codex-skills and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Codex Skills does
Five Codex CLI skills that give OpenAI Codex a Claude Code-style workflow (feature dev, PR review, browser testing, data engineering) plus an orchestrator that routes each task to official Codex plugins first. Workflow rewrites only, with no copied plugin code.
Alternatives in Testing
- Playwright — claude-plugins-official Browser automation, E2E testing, screenshots 29.4k ★
- Pua — Use when the user invokes /pua or asks for PUA mode, try-harder/retry help, change-approach coaching, completi 19.5k ★
- Claude Code Spec Workflow — Automated Kiro-style Spec workflow for Claude Code 3.6k ★
README
Codex Skills: Claude-Style Workflows for OpenAI Codex
Five personal Codex skills that give the Codex CLI the same workflow shape as a tuned Claude Code setup (feature dev, PR review, browser testing, data engineering) plus an orchestrator that routes each task to official plugins first.
     
🧭 Part of **Abdul Raqeeb Khatri's portfolio**: [📂 Hub](https://github.com/Ark310/portfolio) · [🌐 Site](https://ark310.github.io) · [💼 Experience](https://github.com/Ark310/experience)
Overview
I use Claude Code and OpenAI Codex side by side. Claude Code had a mature setup: plugins such as feature-dev, pr-review-toolkit, playwright and data-engineering, plus habits like brainstorm-then-plan, parallel agents and verify-before-done. Codex had none of that. The goal was for Codex to work *like* that setup, using **official OpenAI-curated plugins only** and no copied third-party plugin code.
So these are **workflow-compatible rewrites**, not ports. Each skill describes the *process* (when to explore, when to ask, when to fan out to agents, how to report) and delegates the actual capability to official Codex plugins (`superpowers`, `github`, `coderabbit`, `codex-security`, `build-web-data-visualization`) or to repo-native tooling.
💡 Why I Built It
Claude Code had a mature workflow setup and Codex had none. The goal was for Codex to work the same way using only official OpenAI-curated plugins, with no copied third-party code.
📈 I
Related Skills
Brand Auto
Brand-auto — bounded autonomous brand work. Brand-orchestrator routes to Durham specialists (Aaker/Kotler/Glas
Skill Creator Plus
Based on Anthropic's skill-creator — with bug fixes, working Cowork support, and official best practices baked
QA Learn Conventions
Learn this repo's automation house style (core harness, base classes, page objects, API clients, data, naming)
Spec Orchestrator
Orchestrate implementation of a large specification using parallel agent teams. Reads specs via discovery suba
Zeta
A retro-style TUI File explorer, remember Norton Commander and the like? well, this is my take in Rust while t
Mind World Map
Agent skill for engineering mind maps — closure standard a cold reader can verify, two generations of machine
Related Agents
Harness Eng Lead
Engineering lead — routes each task to one of five specialists by consult-when, owns architecture review for i
Codex Orchestrator
CLI tool for delegating tasks to GPT Codex agents via tmux sessions. Designed for Claude Code orchestration wi
Codex Debugger
Root-cause a failing test, crash, stack trace, or misbehaving feature by delegating the investigation to OpenA