Supabase provides the database and authentication. Cencori provides the memory and agent orchestration. Together, they form a complete open-source equivalent to Firebase + OpenAI.

Why use them together?

Unified Auth: Cencori respects Supabase Auth RLS policies.
Vector Sync: Automatically sync Supabase table rows to Cencori's Vector Store.
Edge Speed: Call Cencori from Supabase Edge Functions with zero cold starts.

1. Setup

Get your connection details:

Codetext

# .env.local
NEXT_PUBLIC_SUPABASE_URL=your-project-url
NEXT_PUBLIC_SUPABASE_ANON_KEY=your-anon-key
CENCORI_API_KEY=cn_...

2. Edge Functions (Deno)

Run Cencori logic directly on the edge. This is perfect for high-performance AI chat endpoints.

Codetext

// supabase/functions/chat/index.ts
import { serve } from "https://deno.land/std@0.168.0/http/server.ts"
 
serve(async (req) => {
  // 1. Get User from Supabase Auth
  const authHeader = req.headers.get('Authorization')
  if (!authHeader) return new Response('Unauthorized', { status: 401 })
 
  // 2. Call Cencori API (Native Fetch)
  const response = await fetch('https://cencori.com/api/v1/chat/completions', {
    method: 'POST',
    headers: {
      'Authorization': `Bearer ${Deno.env.get('CENCORI_API_KEY')}`,
      'Content-Type': 'application/json',
    },
    body: JSON.stringify({
      model: 'gpt-4o',
      messages: await req.json(),
    }),
  })
 
  // 3. Return stream
  return new Response(response.body, {
    headers: { 'Content-Type': 'text/event-stream' }
  })
})

3. Database Triggers (Vector Store)

Automatically embed and index data when you insert it into Supabase.

Enable pg_net extension in Supabase.
Create a trigger to call Cencori.

Codetext

/**
 * TRIGGER: Sync documents to Cencori Vector Store
 */
create trigger on_document_created
  after insert on public.documents
  for each row execute function
  supabase_functions.http_request(
    'https://cencori.com/api/v1/memory',
    'POST',
    jsonb_build_object(
      'content', new.content,
      'metadata', jsonb_build_object('id', new.id)
    )
  );

Now, whenever you add a row to documents, Cencori automatically embeds it and makes it searchable via semantic search.

4. Auth & User Identity

Pass the Supabase User ID to Cencori to isolate memory per user.

Codetext

// Client-side (Next.js)
const { data: { user } } = await supabase.auth.getUser();
 
await cencori.chat.create({
  messages: [...],
  user: user.id // <--- Links memory to this specific user
});