Oh My CodeX
Multi-agent orchestration for OpenAI Codex CLI. Zero learning curve.
Multi-agent orchestration for OpenAI Codex CLI. Zero learning curve.
Don't learn Codex prompts. Just use OMX.
Oh My CodeX (OMX) is a workflow layer for OpenAI Codex CLI. It keeps Codex as the coding engine and adds a docs-shaped path for turning a vague task into clarified requirements, an approved plan, implementation, and verification.
Quickstart
Install Codex CLI
npm i -g @openai/codex
codex loginInstall and set up OMX
npm i -g oh-my-codex
omx setupVerify the install
omx doctor
codex login status
omx exec --skip-git-repo-check -C . "Reply with exactly OMX-EXEC-OK"Send your first task
omx --madmax --highThen run the default guided flow inside the session:
$deep-interview "clarify the authentication change"
$ralplan "approve the safest implementation path"
$ultragoal "turn the approved plan into durable goals and complete them"Docs map
Start with the pages in this order if you are new to OMX.
Quickstart
Install Codex CLI and OMX, choose a setup path, verify the runtime, and send the first task.
Use cases
Pick the right flow for prototypes, plan-first work, durable goals, parallel execution, or QA loops.
Core concepts
Understand agents, skills, hooks, state, and how the pieces fit together.
Migrate
Move from older OMX layouts or related agent setups into the current Codex-native structure.
Using OMX
After the quickstart, these sections work like the product manual.
Agents
Role-specific prompt surfaces for discovery, planning, implementation, review, and verification.
Skills
Workflow and utility entrypoints such as $deep-interview, $ralplan, $ultragoal, and $team.
Hooks
Codex lifecycle integrations, keyword routing, stop gates, and runtime behavior.
Tools
State, HUD, Wiki, Visual, provider bridges, and other runtime capability surfaces.
Integrations
OpenClaw, Clawhip, MCP, and cross-CLI interoperability.
Configure, automate, and learn
Use these pages when you need exact names, operational details, or release history.