Mysql tool and chat agent only returns first row of many

Describe the problem/error/question

I am working with a client troubleshooting an automation. they have a ai chat agent set up and that will lookup information from a mysql database. the database can return up to 50 rows of data, however, the ai agent only returns the first row of the database as a chat response.

query example: “list the names of 5 players who are graduating in 2026”.

the goal is the have the chat model return all the results from their query, not just the first result.

I set up a similar workflow, replaced mysql with an airtable table, and the 5 names were returned, instead of the 1 name when I had mysql connected.

What is the error message (if any)?

no error, only get the first row of data from many rows.

Please share your workflow

Share the output returned by the last node

it’s a chat trigger.

Information on your n8n setup

  • n8n version: cloud
  • Database (default: SQLite):: external mysql
  • n8n EXECUTIONS_PROCESS setting (default: own, main): default
  • Running n8n via (Docker, npm, n8n cloud, desktop app): cloud
  • Operating system: testing in windows and linux

This is a known limitation with the MySQL tool in AI Agent workflows. Based on the search results, there’s a similar issue reported where [github.com]( AI Agent node with batch processing always uses the output of the first tool call · Issue #19592 · n8n-io/n8n · GitHub ) shows AI Agent nodes with batch processing only use the output of the first tool call.

The issue appears to be related to how the AI Agent processes MySQL tool results. Try these solutions:

• Check if there’s a `top_k` or limit parameter in your MySQL tool configuration that might be restricting results to 1 row

• As a workaround, consider using the “Call Another Workflow” tool instead of the direct MySQL tool - create a separate workflow that handles the MySQL query and returns all results

• Test with a simple HTTP Request tool or other data source to confirm if this is MySQL-specific

The fact that Airtable works correctly suggests this is specifically a MySQL tool limitation rather than a general AI Agent issue. You might want to report this as a bug to the n8n team if the workarounds don’t solve it.