Skip to main content
The Phaseo CLI can launch oh-my-pi with a dedicated gateway key and a temporary OpenAI-compatible provider catalog.

​
Set up with the Phaseo CLI

The first interactive run offers to install the omp command when it is missing. Use --catalog all to expose every active Phaseo text/chat model, or --dry-run --json to preview the install and launch plan without creating a key or starting oh-my-pi:
Use phaseo run omp ... for the explicit command form. Put oh-my-pi flags after -- so Phaseo can consume its own flags first.

​
Configure oh-my-pi manually

Use a persistent provider when you want Phaseo to remain available in every oh-my-pi session. Create a Phaseo key, export it for the shell that starts oh-my-pi, and add a provider to ~/.omp/agent/models.yml:
~/.omp/agent/models.yml
Start oh-my-pi with the provider-qualified model id:
Add more model entries to the same provider as needed. Keep apiKey pointed at an environment variable instead of writing the key directly into models.yml.

​
Verify

Send a small prompt through the launched harness, then confirm the request appears in Phaseo activity. The runner writes models.yml under a temporary PI_CODING_AGENT_DIR and removes it when oh-my-pi exits.

​
Troubleshooting

​
oh-my-pi is not installed

Run the command in an interactive terminal to approve the install, or install oh-my-pi separately and rerun with --skip-install.

​
The model is not listed

Use --catalog all or pass a model id with --model. Model discovery is not required for the Phaseo gateway.
Last modified on September 10, 2026