this is f**king insane
someone found a way to use fable 5.1 with free gpt 5.6 luna subagents and never hit usage limits
[here is how to set it up in 3 mins]
1. install 'fable-advisor' repo
2. type /model fable
3. done save this and give it to your agent