So a colleague made a Lambda in Node.js 12.x which allows you to read and write data with vocal requests to some devices that operate in MQTT and which connects to AWS services

He gave me the code, and I was forced to migrate from Node.js 12.x to Node.js 20.x
I updated the dependencies (actions-on-google, Alexa-SDK, Ask-SDK, AWS-SDK, JSONWEBTOKEN, JWK-TO- PEM, UUID) which caused some problems

I was able to correct the problems related to the discovery reading and writing of the MQTT devices, except when it’s done with vocal requests

When I make a vocal request for this product, Alexa says, “hum, I have no answer to that.”
Or sometimes there is just a noise from Alexa, and in the chat it is written “<Audio only Response>”

I tried to use console.log() to understand what’s going on, but I have no return in Amazon Cloudwatch

My Lambda function is based on this sample : sample

I used some advices from this documentation too : docs

I tried some of the solutions presented here : debug-your-smart-home-ckill (but I couldn’t find anything)

Do you know a way to precisely troubleshoot why Alexa answers “hum, I have no answer to that.” or “<Audio only Response>” ?

Thank you