Supabase for Openclaw

A comprehensive toolkit for Supabase database management and vector similarity search within Openclaw Skills.

lucassynnott
v1.0.0
Jan 14, 2026
17
10.4k
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install supabase

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install supabase using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is Supabase?

The Supabase skill provides a powerful bridge between AI agents and your Supabase backend, enabling seamless database operations, vector similarity searches, and table management. It allows agents to execute raw SQL queries, perform CRUD operations with granular filters, and interact with the pgvector extension for advanced AI-driven search capabilities.

By integrating this tool into your Openclaw Skills library, you empower your agents to manage persistent storage, handle complex relational data, and implement Retrieval-Augmented Generation (RAG) workflows directly within their execution environment. The skill is designed to handle both standard PostgreSQL tasks and specialized AI workloads like embedding storage and similarity matching.

Supabase Use Cases

  • Storing and retrieving application state or user data in Supabase tables.
  • Implementing vector-based long-term memory for AI agents using pgvector.
  • Automating database maintenance tasks like deleting expired sessions or updating record statuses.
  • Performing semantic search across documents or knowledge bases stored in a Postgres database.
  • Inspecting table schemas to help agents understand and map data structures dynamically.

How Supabase Works

  1. The AI agent identifies a database-related task, such as saving information or looking up relevant context via embeddings.
  2. The skill invokes the supabase.sh script using the provided environment variables for authentication (URL and Service Key).
  3. For standard queries, the skill uses the PostgREST API or direct SQL execution to interact with the database.
  4. For similarity searches, the skill generates embeddings (optionally via OpenAI) and calls a designated Remote Procedure Call (RPC) function in Supabase.
  5. The results are parsed and returned to the agent in a structured format, allowing it to continue its workflow with real-time data.

Supabase Setup

To use this skill within Openclaw Skills, you must configure your Supabase credentials in your environment:

# Required credentials
export SUPABASE_URL="https://yourproject.supabase.co"
export SUPABASE_SERVICE_KEY="your-service-role-key"

# Optional: OpenAI key for automatic embedding generation
export OPENAI_API_KEY="sk-xxxx"

# Optional: Management API access
export SUPABASE_ACCESS_TOKEN="sbp_xxxxx"

Supabase Data Schema & Taxonomy

The skill interacts with your existing Supabase schema but works best with the following structure for vector operations:

Object Type Description
id bigserial Unique identifier for table records.
content text The primary text content to be searched or stored.
embedding vector(1536) The vector representation of the content (OpenAI standard).
metadata jsonb Flexible JSON object for storing additional context or tags.
RPC Function function A PostgreSQL function (e.g., match_documents) for similarity matching.

Supabase Advanced Features

  • Full pgvector support for high-performance vector similarity searches.
  • Direct SQL execution via the query command for complex database migrations or analytics.
  • Support for Remote Procedure Calls (RPC) to trigger server-side logic and stored procedures.
  • Automated embedding generation using OpenAI's text-embedding-ada-002 model.
  • Built-in schema inspection with the describe command to facilitate agent understanding of data models.
  • Service role authentication to bypass Row Level Security (RLS) for administrative workflows.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*