Configuration
Install flags, Codex target, zero-config defaults, doctor, hooks, and provider/model setup.
LazyCodex keeps configuration minimal. The parts you directly handle are install flags and Codex plugin configuration. Provider, model, routing, hooks, Skills, and Agent Roles are handled by the OmO-based setup underneath.
Zero-config by Default
LazyCodex ships with reasonable defaults and works right after installation. Touch configuration only when defaults do not fit the repository. There is no separate config file you must prepare first.
npx lazycodex-ai installNo global install. Always invoke through npx. This is shorthand for npx --yes --package oh-my-openagent omo install --platform=codex.
This is the Codex light setup: Agent Roles, Commands, hooks, Skills, model routing, and verification defaults.
Codex Target
LazyCodex always targets Codex. The --platform=codex argument is built into the install path of the lazycodex-ai bin, so the harness connects to OpenAI Codex CLI, not another platform. You do not pass --platform yourself.
Prerequisites
- Node.js: any maintained LTS.
npxships with npm, and Bun is not required. - OpenAI Codex CLI or the Codex app, signed in.
Hooks & Lifecycle
Hooks do not run before approval. On the first run after installation, Codex startup review asks you to approve the omo hooks. After upgrades, plugin files changed and no longer match the previous trust hashes, so hooks appear as Modified. Approve them again so the next session can bootstrap the new version.
Install Flags
The default installer is interactive (TUI). It detects subscriptions, helps choose models, and guides provider authentication.
npx lazycodex-ai installThe recommended path is to let an LLM agent run the install and setup: it handles subscription detection, model selection, and provider authentication for you.
For a fully autonomous, no-prompt setup, pass both flags:
npx lazycodex-ai install --no-tui --codex-autonomous--no-tui --codex-autonomous are forwarded together for fully autonomous setup. The lazycodex-ai bin does not interpret these flags; every flag after install is passed through to omo install, which gives them meaning.
Provider & Model Settings
Provider and model setup is owned by OmO, not LazyCodex. During installation, OmO aligns model profiles from Codex configuration and the bundled model-catalog.json; this is the model routing layer.
- The installer connects provider authentication. Letting an agent run it is the recommended path.
- Provider keys are read from the environment. Treat all
*_API_KEYand OAuth credentials as secrets; never log or commit them. - Deeper provider/model tuning beyond installation follows OmO conventions. Refer to OmO docs for provider env vars and model interpretation rules.
Do not invent provider keys. Supply the keys documented by the provider you chose, and let the installer interpret routing.
Doctor
If anything looks pending or degraded, run:
npx lazycodex-ai doctorIt checks plugin cache, hooks, MCP Servers, Agent Roles, and Codex config state in a health report. It explains what is misconfigured, why, and which field or component to fix.