Supabase Vector Store Extended icon

Supabase Vector Store Extended

Work with your data in Supabase Vector Store with Schema and Namespace support

Actions4

Retrieve

AI-generated

Overview

This node, named "Supabase Vector Store Extended," is designed to interact with a Supabase vector store database. It supports operations such as loading, inserting, retrieving, and updating documents within the vector store. The node extends basic functionality by allowing users to specify a custom database schema and use namespaces for logical partitioning of documents.

The Retrieve operation (the focus here) is intended to fetch similar documents from the vector store based on a query. This can be useful in scenarios like semantic search, recommendation systems, or any application requiring similarity-based document retrieval.

Practical examples:

  • Retrieving documents similar to a user query in a knowledge base.
  • Fetching related product descriptions in an e-commerce platform.
  • Finding contextually relevant text snippets for chatbot responses.

However, note that the actual retrieval logic is not implemented yet; the node currently returns a placeholder message.

Properties

Name Meaning
Use Custom Schema Boolean flag to decide whether to use a database schema other than the default "public" schema.
Schema Name of the database schema to use if "Use Custom Schema" is enabled. Defaults to "public".
Table Name Name of the table in the Supabase database where documents are stored. Required field.
Namespace Logical partition identifier for documents, used to filter data by namespace column.
Options Collection of additional options:
- Query Name Name of the query to use for matching documents. Default is "match_documents".
- Clear Namespace (Only for some operations) Boolean indicating whether to clear the namespace before inserting new data.

Output

The output JSON contains a simple structure indicating the operation performed and a placeholder message since the actual retrieval logic is not implemented. For example:

{
  "operation": "retrieve",
  "message": "Retrieve operation not yet implemented",
  "itemIndex": 0
}

No binary data output is produced by this node.

Dependencies

  • Requires a configured Supabase API key credential to connect to the Supabase service.
  • No external libraries beyond those bundled with n8n and the standard environment appear necessary.
  • The node expects access to a Supabase database with vector store tables.

Troubleshooting

  • Operation Not Supported Error: If an unsupported operation is specified, the node throws an error stating the operation is not supported.
  • Not Implemented Messages: Currently, all operations including Retrieve return a "not yet implemented" message. Users should expect no functional behavior until the implementation is completed.
  • Missing Required Fields: Ensure the "Table Name" property is provided; otherwise, the node may fail or behave unexpectedly.
  • Credential Issues: Make sure the Supabase API key credential is correctly set up and has appropriate permissions.

Links and References

Discussion