SuperAPI AI Directory

deploying-supabase-application

Deployment & operations

View repository

Updated 10-15 20:11

About this skill

Deploying a Supabase application

Overview

Deploy backend changes with Supabase GitHub integration: connect Git in the dashboard, then merge to the production branch to apply migrations and deploy declared Edge Functions. No supabase link or manual db push for the default workflow.

Deployed on merge to production branch Not deployed automatically
New files in migrations/ Seed files (seed.sql, etc.)
Edge Functions declared in config.toml Undeclared functions
Storage buckets declared in config.toml Edge Function secrets; Auth, SMTP, Site URL, API settings in config.toml

Configure secrets and auth in the Supabase dashboard after deploy.

Agent workflow

Follow this skill when deploying a Supabase app or connecting it to Supabase GitHub integration. Run these steps in order. Works wherever the agent can read and write repository files.

Steps

  1. Discovery — Locate the Supabase root (config.toml + migrations/), dashboard working directory, production branch, declared Edge Functions, and storage buckets. Ask the user when multiple candidates exist. Default production branch to main only when repo convention is obvious. Follow discovering-supabase-root.md.

  2. Validate locally — From the Supabase root, before first connect:

    supabase db start
    supabase db reset --local   # or: supabase migration up --local
    supabase db lint
    

    If Edge Functions exist, run project-specific checks (e.g. deno check, deno test under functions/). Fix errors before connecting GitHub or merging to the production branch.

  3. Document dashboard settings — User connects Git in Project Settings → Integrations → GitHub. Provide a filled-in table from configuring-github-integration.md:

    Setting Value
    Git repository User selects repo
    Working directory Discovered Supabase root path
    Production branch User's branch
    Deploy to production Enabled
  4. Post-connect verification — After the user enables integration and merges to the production branch:

    • Confirm deploy succeeded in the Supabase dashboard
    • Run project-specific smoke checks for schema or function behavior changes

    First merge to the production branch populates a new remote project. Troubleshooting: troubleshooting-deployment.md.

Decision tree

Deploy Supabase backend?
├─ Locate config.toml + migrations/ → working directory for dashboard
├─ Migrations apply + db lint pass locally?
│  └─ No → fix before connect or merge
├─ User connects GitHub integration + enables Deploy to production
└─ Merge to production branch → verify in dashboard

Reference index

Doc When to use
discovering-supabase-root.md Supabase root, working directory, local validation
configuring-github-integration.md Dashboard GitHub integration settings
troubleshooting-deployment.md Local validation, deploy, and integration failures