Spaces:
Running
Running
Supabase Edge Functions
This directory contains Edge Functions that can be deployed to Supabase.
Available Functions
urdf-parser
: Extracts metadata from Urdf files using OpenAIcreate-animation
: Generates animations for Urdf models based on user instructions
Deployment Instructions
Prerequisites
Install Supabase CLI:
npm install -g supabase
Login to Supabase:
npx supabase login
Set your Supabase project ID:
npx supabase link --project-ref mizajlqhooderueazvnp
Deploy a Function
To deploy a function, run:
node deploy-function.js urdf-parser
Or use the Supabase CLI directly:
npx supabase functions deploy urdf-parser
Environment Variables
The edge functions require environment variables to be set in your Supabase project:
OPENAI_API_KEY
: Your OpenAI API key for the AI functionality
To set these variables, use:
npx supabase secrets set OPENAI_API_KEY=your-api-key
Testing Deployed Functions
After deployment, you can test your function with:
curl -X POST 'https://mizajlqhooderueazvnp.supabase.co/functions/v1/urdf-parser' \
-H 'Authorization: Bearer eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpc3MiOiJzdXBhYmFzZSIsInJlZiI6Im1pemFqbHFob29kZXJ1ZWF6dm5wIiwicm9sZSI6ImFub24iLCJpYXQiOjE3NDE4NzU1MTgsImV4cCI6MjA1NzQ1MTUxOH0.iTbP3KqPr4Fl48YQnlwplyPF3OCTjU5g62WmmygjSbA' \
-H 'Content-Type: application/json' \
--data '{"urdfContent":"<robot name=\"test\">...</robot>"}'