Ai agent with Supabase Vector Store have issue. Its always refer the documents table?
My transcripts table is empty. but its giving the result from documents table.
If I rename the documents table, Its giving error about documents table. Is documents table are fixed for As tools for AI Agent?
HI @mashrmail
You’re not locked to a documents table in n8n’s Supabase Vector Store—just set the Table Name to whatever you need (e.g., transcripts). If your AI Agent keeps querying documents, it’s because that field is still set to “documents,” so update it in the node and any related agent tool configs. Also make sure the target table exists with the expected schema (embeddings, content, metadata). If you see errors after renaming, double-check those settings
Table name is drop down in n8n. I have selected correct one that is ‘users’ Now i have created a vector table ‘users‘ with empty record. You may see result are still fetching from documents.
I strongly feel in the backend on n8n when supabase vector called in the given situation, Its fixed by n8n developer to call documents table.
Example: select * from documents.
I have seen it’s bug its always call match_documents function to supabase vector store DB.
It should call dynamic according to table name: match_documents , match_users, match_transcripts…and so on
