Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
Connects to a Supabase project to run SQL queries, perform CRUD operations on tables, and execute vector similarity search using pgvector. The service role key grants full database access; an anon key can be used for restricted access.
Handles raw SQL, CRUD, and vector similarity search through one CLI without needing separate database clients or a custom embedding pipeline.
When to use it
- Query user records filtered by status or date range
- Insert rows collected from a form or API response
- Find semantically similar documents using pgvector embeddings
- Inspect table schemas before writing migration SQL
- Delete expired session rows matching a condition
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Find the top 5 documents related to 'setting up authentication'
- 1Calls vector-search on the documents table with the query string
- 2Generates an embedding for the query using OpenAI text-embedding-ada-002
- 3Passes the embedding to the match_documents RPC function in Supabase
- 4Retrieves rows ranked by cosine similarity above the configured threshold
- 5Returns the top 5 matching records with content and similarity scores
A ranked list of 5 document records with content fields and similarity scores
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.