I have identified a potential regression within the Python SDK when attempting to programmatically update custom slot types in an Architect bot flow. Specifically, the put_architect_bot_slot_type method consistently fails with a 400 Bad Request error when the list of synonyms contains non-ASCII characters, despite the payload being valid UTF-8. I have verified the behavior with raw API calls and the issue seems isolated to the SDK’s serialization logic. Has anyone else encountered this behavior when managing multi-language bot configurations via the SDK?
Hello. I have seen this when I export recording metadata for our legal team in different regions. The SDK sometimes has problems with the encoding of special characters.
I found that I had to manually encode the string to UTF-8 before I passed it to the SDK function. It is very annoying when the SDK says it is supported but it fails during the transmission.
You should try to use the raw request utility in the SDK to see if that works better for your Japanese bot.
Hey. I am an admin for our Edge servers and I do not do much bot coding, but I have seen our devs complain about the Python SDK a lot. They usually just give up and use the requests library directly when the SDK starts acting weird.
It is a lot easier than trying to figure out why the official tool is broken. Maybe you should just post the bug on the Genesys GitHub page? They are usually pretty good at fixing the SDK if you give them a clear example.