Skip to content

AI Agent node common issues#

Here are some common errors and issues with the AI Agent node and steps to resolve or troubleshoot them.

Internal error: 400 Invalid value for 'content'#

A full error message might look like this:

1
2
3
Internal error
Error: 400 Invalid value for 'content': expected a string, got null.
<stack-trace>

This error can occur if the Prompt input contains a null value.

You might see this in one of two scenarios:

  1. When you've set the Prompt to Define below and have an expression in your Text that isn't generating a value.
    • To resolve, make sure your expressions reference valid fields and that they resolve to valid input rather than null.
  2. When you've set the Prompt to Take from previous node automatically and the incoming data has null values.
    • To resolve, remove any null values from the chatInput field of the input node.

Error in sub-node Window Buffer Memory#

This error displays when n8n runs into an issue with the Window Buffer Memory sub-node.

It most often occurs when your workflow or the workflow template you copied uses an older version of the Window Buffer Memory node.

Try removing the Window Buffer Memory node from your workflow and re-adding it, which will guarantee you're using the latest version of the node.

A Chat Model sub-node must be connected error#

This error displays when n8n tries to execute the node without having a Chat Model connected.

To resolve this, click the + Chat Model button at the bottom of your screen when the node is open, or click the Chat Model + connector when the node is closed. n8n will then open a selection of possible Chat Models to pick from.

No prompt specified error#

This error occurs when the agent expects to get the prompt from the previous node automatically. Typically, this happens when you're using the Chat Trigger Node.

To resolve this issue, find the Prompt parameter of the AI Agent node and change it from Take from previous node automatically to Define below. This allows you to manually build your prompt by referencing output data from other nodes or by adding static text.