Documentation

Learn how to install, configure, and use LazyCodex for ultrawork.

Overview

LazyCodex is the lazy way to get oh-my-openagent (OmO) up and running inside Codex. Think LazyVim for lazy.nvim, but for Codex.

What you get

OmO is a full AI dev team in a harness. It gives you discipline agents (Sisyphus orchestrates Hephaestus, Oracle, and Librarian), parallel execution, multi-model routing, a skills system, hooks and lifecycle, and zero-config defaults. LazyCodex wraps the entire setup so you don't have to think about it.

The one-word workflow

You don't need to ultrathink. Just prompt with {your prompt} ultrawork. One word, and every agent activates.

How it fits together

LazyCodex is a thin distribution layer. The core engine is OmO. LazyCodex is maintained by Sisyphus Labs.

Installation

One command gets you running. LazyCodex is currently available on OpenCode; the Codex edition is coming June 2026.

Install

bunx lazycodex install

This is exactly equivalent to bunx omo install --platform=codex.

Autonomous one-liner

bunx lazycodex install --no-tui --codex-autonomous

Prerequisites

Do NOT use npm install -g or bun add -g. Always invoke via bunx.

Let an agent do it

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.

ultrawork mode

ultrawork is the headline mode. Typing ultrawork (or the short alias ulw) anywhere in your prompt activates maximum-precision, outcome-first, evidence-driven orchestration.

"One word. Every agent activates. Doesn't stop until done."

Usage

ulw add authentication

What it enforces

  • Strict TDD: RED → GREEN → SURFACE → CLEAN
  • At least 3 realistic QA scenarios
  • Real manual-QA channels (HTTP call, tmux, browser)
  • A binding verification gate that loops until the work is genuinely done

$ulw-loop

$ulw-loop is a self-referential development loop that runs until verified completion.

How it works

The agent works continuously and emits <promise>DONE</promise> when it believes the task is complete, but that does NOT end the loop. An Oracle must verify the result first. The loop ends only after the system confirms Oracle verified it. If verification fails, it continues with the message: "Oracle verification failed. Continuing ULTRAWORK loop."

Syntax

/ulw-loop "task description" [--completion-promise=TEXT] [--strategy=reset|continue]

Limits

The iteration cap is 500 in ultrawork mode (100 in normal mode).

$ulw-plan

$ulw-plan is the strategic planning consultant (Prometheus). It turns an idea into a decision-complete work plan. It is a planner, NOT an implementer. When you say "do X" it produces a plan for X and never writes product code.

The flow

  1. Socratic interview
  2. Parallel codebase exploration
  3. Metis gap analysis
  4. Writes the plan to plans/<slug>.md
  5. Optional Momus high-accuracy review

Output

Questions, research, and a work plan you can hand to $start-work.

$start-work

$start-work executes a Prometheus work plan until every top-level checkbox is done.

How it works

  • Durable Boulder state in .omo/boulder.json survives across turns and sessions
  • A Stop-hook re-injects the next turn until the plan is complete
  • Independent sub-tasks fan out to parallel subagents
  • Strict TDD plus five evidence gates: plan reread, automated verification, manual-QA, adversarial QA, cleanup
  • Progress is recorded to a ledger

Syntax

/start-work [plan-name] [--worktree <absolute-path>]

Done

It prints an ORCHESTRATION COMPLETE block when every checkbox is checked.