Supabase Database Toolkit - OpenClaw Skill

Full Supabase database toolkit - schema explorer, natural language queries, CRUD, pgvector search, RLS policies, migrations, health checks, data export, Warehouse analytics (pg_duckdb), automatic embeddings, queues (pgmq), Realtime, OAuth 2.1, Storage CDN, Edge Functions, and integrations.

What it does

• Schema explorer - list tables, describe columns/types/relationships, view RLS policies and indexes
• Natural language queries - describe what you want in plain English, get SQL results
• CRUD operations - insert, select, update, upsert, delete with powerful filters
• Vector search - similarity search with pgvector embeddings
• RLS policy helper - generate Row Level Security policies from natural language descriptions
• Migration management - create, apply, and check status of schema migrations
• Health checks - connection status, database size, active connections, slow queries, index usage
• Data export - export tables as CSV or JSON with filters
• Warehouse / pg_duckdb - 600x faster analytics queries, Iceberg-format Analytics Buckets on S3
• Automatic embeddings - trigger-based pipeline auto-generates vector embeddings on insert/update
• Queues / pgmq - exactly-once message delivery with configurable visibility windows
• Realtime - Postgres Changes, Broadcast, Presence, and Broadcast from Database
• Auth - OAuth 2.1 server mode, X/Twitter OAuth 2.0
• Storage - 14.8x faster listing, Smart CDN, image transformations, resumable uploads, S3 compatibility
• Edge Functions - regional invocations, NPM compatibility, WASM support
• Integrations - Claude Connector (MCP), Stripe Sync Engine, Log Drains, PrivateLink
• Infrastructure - Database Branching, Read Replicas, Supabase Vault

Quick start

Install the skill

git clone https://github.com/mvanhorn/clawdbot-skill-supabase.git ~/.openclaw/skills/supabase

Set up credentials

export SUPABASE_URL="https://yourproject.supabase.co"
export SUPABASE_SERVICE_KEY="eyJhbGciOiJIUzI1NiIs..."

# Or use the new project-scoped key (March 2026+)
export SUPABASE_API_KEY="sb_secret_..."

# Optional: direct Postgres access (enables migrations, health checks, schema dump)
export SUPABASE_DB_URL="postgresql://postgres:[PASSWORD]@db.[REF].supabase.co:5432/postgres"

Example chat usage

• "Show me all the tables in my database"
• "Describe the schema for the orders table"
• "Query users who signed up in the last 7 days"
• "Insert a new product with name 'Widget' and price 29.99"
• "Search my documents for anything about authentication"
• "Generate an RLS policy so users can only see their own data"
• "Check database health - connections, size, slow queries"
• "Export the orders table as CSV"
• "Create a migration to add a comments table"
• "Run an analytics query on pageviews from the last 30 days"
• "Set up automatic embeddings on my documents table"
• "Send a message to my task queue"
• "Show me who is online via Realtime Presence"

Commands

| Command | What it does | |---------|-------------| | query | Run raw SQL (including Warehouse/pg_duckdb queries) | | select | Query with filters (--eq, --gt, --lt, --like, --limit, --order, etc.) | | insert | Insert one or more rows | | update | Update rows matching filters | | upsert | Insert or update | | delete | Delete rows matching filters | | vector-search | Similarity search via pgvector | | tables | List all tables with row counts | | describe | Show table schema (columns, types, constraints) | | schema | Full schema dump (requires SUPABASE_DB_URL) | | rpc | Call a stored procedure | | health | Database health check | | export | Export data as CSV or JSON | | migration | Create, apply, or check migration status |

How it works

Wraps the Supabase REST API (PostgREST) with a shell script CLI. Uses the service role key for full access (bypasses RLS). Direct Postgres features (migrations, health, schema dump) require SUPABASE_DB_URL. Vector search requires pgvector extension and OPENAI_API_KEY for embedding generation. Warehouse queries run through pg_duckdb for columnar analytics acceleration.

Environment variables

| Variable | Required | Description | |----------|----------|-------------| | SUPABASE_URL | Yes | Project URL | | SUPABASE_SERVICE_KEY | Yes* | Service role key (legacy) | | SUPABASE_API_KEY | No* | Project-scoped key (new, March 2026+) | | SUPABASE_DB_URL | No | Direct Postgres connection string | | OPENAI_API_KEY | No | For generating embeddings | | SUPABASE_ACCESS_TOKEN | No | Management API token |

*Either SUPABASE_SERVICE_KEY or SUPABASE_API_KEY required.

What's new in v3.0.0

• Warehouse / pg_duckdb - 600x analytics acceleration, Iceberg Analytics Buckets on S3
• Automatic embeddings - trigger-based pipeline with pgmq + pg_cron + Edge Functions
• Queues / pgmq - exactly-once delivery, pgmq_public for client consumers, worker patterns
• Realtime - Postgres Changes, Broadcast, Presence, Broadcast from Database (beta)
• OAuth 2.1 Server - Supabase as identity provider (Nov 2025 beta)
• X/Twitter OAuth 2.0 - replacing legacy 1.0a flow (Feb 2026)
• Storage improvements - 14.8x faster listing, Smart CDN, image transforms, resumable uploads, S3 compat
• Edge Functions - regional invocations, rate limiting, NPM/WASM support
• Claude Connector - 32 MCP tools for Supabase
• Stripe Sync Engine - query Stripe data via SQL
• Log Drains - Datadog, Grafana, Sentry, Axiom, S3 (Pro plan)
• PrivateLink - AWS VPC connectivity (Jan 2026)
• Database Branching - preview environments with isolated schema copies
• Read Replicas - regional read-only replicas for lower latency
• Supabase Vault - encrypted secrets storage in Postgres
• Project-scoped API keys - sb_publishable_ / sb_secret_ format
• OpenAPI spec restricted - service_role keys only (March 11, 2026)

API key migration (March 2026)

Supabase is deprecating legacy JWT-format keys. New project-scoped keys use sb_publishable_ and sb_secret_ prefixes. Both formats work during the transition (through late 2026). See SKILL.md for the full migration guide.

License

MIT