AI Dashboard developer reference

Build AI-assisted operator workflows while keeping the browser, Ordering.co API, standalone services, channels, providers, and storage systems independently accountable.

Choose the right reference

Goal	Reference
Understand AI agents, provider schemas, conversations, channels, imports, and external integrations	Agents, channels, and provider integrations
Decide whether code belongs in the frontend, shared package, Ordering.co API, FastAPI services, Supabase functions, or storage	Services and data boundaries
Set up the monorepo, run tests, build the app, or prepare an authorized release	AI Dashboard engineering guide
Implement authentication or a server request	Use the canonical API Reference in the documentation header. Start with POST /auth.

The integration and service pages explain responsibilities. The engineering guide explains contributor work. They do not duplicate server schemas, provider contracts, or product instructions.

System shape

The repository is a Yarn and Turbo monorepo with:

• apps/dashboard, the Vite and React operator application;
• packages/shared, shared API, authentication, configuration, state, provider, and UI utilities;
• services/menu-import and services/meta-ads, standalone Python services with independent runtime and secret requirements; and
• supabase/functions, server-side functions for selected link-in-bio and import-log workflows.

The Ordering.co API remains the authorization and core commerce-data boundary. The frontend can present provider-specific forms and status, but it cannot guarantee channel delivery, provider acceptance, or data persistence. Availability depends on the deployed versions and project configuration.

For operator workflows, use the AI Dashboard product documentation.