Getting Started
Install LazyCodex into Codex with one npx line, authenticate, and run your first task with init-deep, ulw-plan, start-work, and ulw-loop.
LazyCodex installs the OmO agent harness into Codex with a single npx line — no global install. Get it set up, then drive your first task.
Prerequisites
- Bun
- The OpenAI Codex CLI
LazyCodex runs inside OpenAI Codex, so the Codex CLI must be installed and working.
Install
One line. No global install, no npm i -g — always use npx:
npx lazycodex-ai installThis is shorthand for the OmO installer targeting the Codex platform:
npx --yes --package oh-my-openagent omo install --platform=codexFor a fully autonomous, no-TUI setup, add the autonomous flags:
npx lazycodex-ai install --no-tui --codex-autonomousThe installer wires the OmO harness — commands, skills, hooks, and model routing — into your Codex setup. It is strongly recommended to let an LLM agent run the install and walk the setup for you; the agent handles subscription detection, model selection, and provider auth automatically.
Authenticate
There is no separate LazyCodex login command. The installer (and any LLM agent you let run it) handles subscription detection, model selection, and provider auth automatically. Being logged into the Codex CLI itself is a prerequisite, but it is not a LazyCodex auth step.
For provider and routing details, see Configuration.
Run your first task
Start by building project memory, then reach for the command pillars. Invoke the pillars with the $command syntax the installer prints.
First, generate hierarchical AGENTS.md project memory so agents start with full context:
/init-deepThen choose a pillar for the work:
$ulw-plan "what to build"— Prometheus plans the approach and writes it toplans/<slug>.md. It never writes product code.$start-work [plan-name] [--worktree <path>]— executes a plan until every checkbox is done, then prints ORCHESTRATION COMPLETE.$ulw-loop "task" [--completion-promise=TEXT] [--strategy=reset|continue]— a self-referential loop that runs until Oracle-verified completion.
A typical flow plans first, then executes:
/init-deep
$ulw-plan "build a REST API for managing tasks"
$start-workOr, for an open-ended task that should keep moving until the result is verified by evidence, use the loop directly:
$ulw-loop "fix the failing checkout flow and verify it end to end"Skills
Beyond the command pillars, LazyCodex installs OmO's skill layer for specialized work — review-work, remove-ai-slops, frontend-ui-ux, programming, LSP, rules, comment-checker, and more. The command layer stays simple while skills add specialist judgment. See Skills for the full library.
Next steps
Commands
The command pillars: ulw-plan, start-work, ulw-loop, init-deep.
Concepts
The ultrawork loop, discipline agents, and model routing.
Use cases & guide
Worked examples for everyday tasks.
- The underlying engine: oh-my-openagent (OmO)
- LazyCodex on GitHub
LazyCodex
The OmO agent harness packaged for Codex — project memory, planning, execution, and verified completion. LazyVim for Codex.
Use Cases & Guide
Pick the right command for the job — plan-first work, verified open-ended loops, codebase onboarding, worktree-isolated execution, and post-implementation review.