Add stream_speech, list_folder_speeches, and fix bug with move_to_trash_bin

[BrandonKates]

– Added stream_speech
– Added list_folder_speeches
– Fixed bug with move_to_trash_bin

Stream speech adds the ability to return the content of the speech directly rather than saving to a file. I considered refactoring download_speech but I felt this was cleaner and ensures backwards compatibility.

List folder speeches enables you to query speeches from only a specific folder. I found that the folder parameter on get_speeches doesn't seem to work, and this is the query that Otter uses when you are looking into a folder via the UI. Since we're trying to mimic the API I didn't want to refactor the get_speeches() function. It also allows for pagination via the last_load_speech_id argument.

Move to trash bin was failing with a referrer error, so I made sure to add the referrer already present in a number of other functions.

[ChocoTonic]

Hey @BrandonKates,

Thanks for these additions! If you'd like to see these changes, feel free to make the updates in this fork at ChocoTonic/otterai-py since it seems like this project is unmaintained. It would be great if each change could be submitted as a separate PR to keep things organized.

Appreciate your work on this!

[ChocoTonic]

Adding a bit more context here. While I do have collaborator access, the repository has been largely untouched and I still do not have permission to manage repo secrets, which is required for the integration tests I am trying to add. It has been over a month with no movement on that front.

Given the lack of maintenance activity and the blocker around basic permissions, I am moving forward with active maintenance in the restarted repository at https://github.com/ChocoTonic/otterai-py/ so development and testing can continue without friction.

Going forward, please open any pull requests or contributions against that repository so changes can be reviewed and merged without delay.