Skip to main content
Full Mode runs your project’s entire front end inside Autonomy while proxying its API calls to your existing staging backend. Instead of editing isolated components, you traverse and edit the whole live app - and its real data comes from your staging environment, not from Autonomy.

​
When to use it

Use Full Mode when the user wants to work on the whole application (navigating real screens, real data) rather than a single component in isolation.

​
Enabling it

1

Finish setup basics

Make sure the project is set up first - load the setup skill if needed (load_autonomy_skill name="setup").
2

Set the dev command + URL

update_setup_config(dev_command, dev_url) - the command and URL that serve the app (e.g. npm run dev, http://localhost:5173).
3

Point the proxy at staging

update_preview_proxy_config(cors_hosts=[<staging API host>], forwarded_origin=...) - this routes the app’s API calls to your staging backend. This is the core of Full Mode. Use forwarded_origin="${PUBLIC_ORIGIN}" if the app’s auth (Clerk / NextAuth) needs it.
4

Add any front-end env

update_setup_files(files=[{ file_path: ".env", file_content: "..." }]) - any env the front end needs, such as the staging API base URL.
5

Get the preview URL

get_setup_preview_status returns the live preview URL of the whole app.
Full Mode is feature-flagged in the platform (the project_cors_proxy flag). If a preview-config call is rejected with 403, the project doesn’t have Full Mode enabled.
  • get_preview_proxy_config / update_preview_proxy_config - the proxy config.
  • get_setup_files / update_setup_files - files written into the sandbox (e.g. .env).
  • get_setup_preview_status - dev-server state + live preview URL.