Warning

This repo is archived. It won't receive new features, bug fixes, or security updates.

Published releases (up to v0.0.6) stay available and will keep working in existing workflows.

---

Supabase Embeddings Generator

A GitHub Action that converts your markdown files into embeddings and stores them in your Postgres/Supabase database, allowing you to perform vector similarity search inside your documentation and website.

This action is a companion to the headless-vector-search repo, which is used to store and retrieve the embeddings using OpenAI and Supabase.

Usage

In your knowledge base repository, create a new action called .github/workflows/generate_embeddings.yml with the following content:

name: 'generate_embeddings'
on: # run on main branch changes
  push:
    branches:
      - main

jobs:
  generate:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v3
      - uses: supabase/embeddings-generator@v0.x.x # Find the latest version in the repo's releases
        with:
          supabase-url: 'https://your-project-ref.supabase.co'
          supabase-service-role-key: ${{ secrets.SUPABASE_SERVICE_ROLE_KEY }}
          openai-key: ${{ secrets.OPENAI_KEY }}
          embedding-model: 'text-embedding-3-large' # optional; defaults to text-embedding-ada-002
          docs-root-path: 'docs' # the path to the root of your md(x) files

Make sure to set SUPABASE_SERVICE_ROLE_KEY, and OPENAI_KEY as repository secrets in your repo settings (settings > secrets > actions).

You can override the OpenAI embedding model by providing the optional embedding-model input (defaults to text-embedding-ada-002 if not supplied).

See the instructions in the headless-vector-search for more information on how to query your database from your website.

Developers

See details in MAINTAINERS.md

License

MIT