I know we can limit the AI usage of our Agents through different types of quotas.
I wonder if it could be good to have a distinction about what models each group can use in the same Agents.
We want to preserve expensive tokens for TL2/3 but also allow a less expensive model for TL1 users.