I’m trying to use an AI Agent node to query a Pinecone vector store. The Run 1 through the Model keeps returning no output I can see in the chat panel or when checking on the AI Agent node.
What is the error message (if any)?
None
Please share your workflow
Share the output returned by the last node
Output appears to fail to search the vector store correctly and always returns a sorry we couldn’t find it style response.
To ensure your data is properly embedded and indexed:
The Pinecone Vector Store node in “Insert Documents” mode sends the text to the index. Then, using “Get Many” mode, you can run manual search tests. If it works here, your embeddings and data are properly indexed. The “Populate a Pinecone vector database from a website” example in the documentation shows how to scrape, embed, and query without agents.
The documentation specifies that the Pinecone Vector Store node should connect directly to the AI Agent’s tools connector using the “Retrieve Documents (As Tool for AI Agent)” mode. In that mode, you can define the name and description, which are visible to the model and tell it when to use the tool.
Before integrating it into the agent, connect the Pinecone Vector Store node to the Retrieve Documents node and run a test query (Prompt + number of results). If you get relevant documents, the vector store is working properly.