Cleans up the user permissions for ai tutor

to do still

• Before merge, audit the list of users here:
  Code.org vs the list of users here Code.org to see if we want to add anyone to the pilot,
  Then revoke the UserPermission from everyone before deleting it, so we don't orphan them.
  Do this on every env?

• replace the tutor2 check with trust_chat_client

• expose has_ai_tutor_access to the frontend (similar to how aichat does it by exposing a controller action?)

• hook it up to the frontend / re think through when to show tutor vs not vs sleeping. is tutor always there all the time but sleeping? do we need a separate check for "enabled by experiment" and "disabled" so we can differentiate between sleeping and not visible? (GA would be always sleeping everywhere unless ai_tutor_available?)

Links

• Jira:

Testing story

Deployment strategy

Follow-up work

Privacy

Security

Caching

PR Creation Checklist:

• Tests provide adequate coverage
• Privacy impacts have been documented
• Security impacts have been documented
• Code is well-commented
• New features are translatable or updates will not break translations
• Relevant documentation has been added or updated
• User impact is well-understood and desirable
• Follow-up work items (including potential tech debt) are tracked and linked