FEATURE: llm quotas

[SamSaffron]

Adds a comprehensive quota management system for LLM models that allows:

• Setting per-group token and usage limits with configurable durations
• Tracking and enforcing token/usage limits across user groups
• Quota reset periods (hourly, daily, weekly, or custom)
• Admin UI for managing quotas with real-time updates
• Full test coverage for quota models and controllers

This system provides granular control over LLM API usage by allowing admins
to define limits on both total tokens and number of requests per group.
Supports multiple concurrent quotas per model and automatically handles
quota resets.

[SaifMurtaza]

Are we able to provide guidelines for setting the max tokens? Some tooltips would be helpful here, i.e - for the non-technical admin,concepts such as tokens might be missed. They are probably left wondering "How many words does this actually mean?" or "How many times can I message the persona?"

[keegangeorge]

LGTM overall, just a few suggestions here and there

[SamSaffron]

OK @keegangeorge I think it is all addressed! thanks for the review

[SamSaffron]

thanks @SaifMurtaza , will add more clarity, I am on the fence on implementing an optional "absolute quota" that is shared on the group... then you can guarantee you would never spend more than N$ on AI a day even if you have very large groups.

[davidtaylorhq]

As mentioned on dev, the use of { i18n } from "discourse-i18n"; should be reverted until after the next core version bump.

[keegangeorge]

LGTM! Just two small fixes to fix linting and .discourse-compatibility needed before merging 🚀

[keegangeorge]

See internal: /t/124366/17 we should pin discourse-ai for core < beta4 in .discourse-compatibility