Investigate OpenAI assistant cancellation reason: max_prompt_tokens

[SmittieC]

See this trace. OpenAI cancelled the run due to max_prompt_tokens. Is this something we can tweak? Maybe we'll have to use another truncation strategy?

We also need to give the user a better error message when this happens. From JJ:

(anytime we're confused like this, lets make part of the resulotion better clarity of the error to the end user)

[snopoke]

It can be changed (https://platform.openai.com/docs/api-reference/runs/createRun#runs-createrun-max_prompt_tokens) though since we don't provide one at all I wonder why it's still happening.

Some more info here: https://platform.openai.com/docs/assistants/deep-dive#max-completion-and-max-prompt-tokens