Continous refresh chat and Improperly formed request.

## Problem

request some task and it responses

**I understand that we're starting a fresh conversation. How can I help you today with your project in the tax-collector workspace?** and after
**Improperly formed request.

Request ID: af99ac23-9e70-4ed4-9b96-25ff089fb7e7**

Logs were

2025-07-27 16:38:36.328 [info] [Error - 4:38:36 PM] [2025-07-27T14:38:36.326Z] lserver: {
"code": "QModelResponse",
"requestId": "f3fab84c-c29f-4400-a20a-aa54dcbb5915",
"name": "Error",
"message": "Improperly formed request.",
"cause": {
"name": "ValidationException",
"$fault": "client",
"$metadata": {
"httpStatusCode": 400,
"requestId": "f3fab84c-c29f-4400-a20a-aa54dcbb5915",
"attempts": 1,
"totalRetryDelay": 0
},
"reason": null
}
}

## Steps to reproduce the issue

just chat, "please resolve the last console errors"

## Expected behavior

execute the code task

## System details (run `AWS: About` and/or `Amazon Q: About`)

2025-07-27 16:38:21.817 [info] [Info - 4:38:21 PM] [2025-07-27T14:38:21.812Z] lserver: System Information: {"languageServerVersion":"1.23.0","clientName":"vscode","clientVersion":"1.102.2","OS":"win32","OSVersion":"10.0.26100","ComputeEnv":"unknown","extensionVersion":"1.85.0"}

Start by reproducing the issue in the VS Code Amazon Q chat flow using the prompt “please resolve the last console errors,” then inspect the provided language-server logs and the AWS: About/Amazon Q: About entry points for version context. Done means the request completes the code task without returning “Improperly formed request.”

Written by the indexing model from the issue text.