Siqi Chen#1537
pro-tip: fable is really good at writing workflows, but very expensive.
let fable orchestrate a fleet of gpt-5.5 agents and get more bang for your buck!
Users praise Fable for being super precise and solid when orchestrating GPT-5.5 agents with good constraints to reduce workflow costs.
@sawyerhood I’ve been playing around with having it orchestrate GLM and Kimi, getting really token efficient
@ToastingTweets this is: https://github.com/SawyerHood/omegacode
@sawyerhood What tool is this?
@sawyerhood You're always the first to ship the stuff we plan to do @niteshiftdev :) At this point I should just set-up an automation to implement all your Github projects as Niteshift features (and credit you at http://niteshift.dev/changelog)
@sawyerhood works with deepseek as well, fable can be super precise on what to build/do and with good constraints v4 is solid
@sawyerhood Yeah, currently i am at the first stage of it. codex babysitting claude so that it wont go on random errands, once I get confident then i could switch it the other way and do nested loops?
codex babysitting claude which manages a fleet of codex lol.
@morganlinton @sawyerhood How do you do that
@sawyerhood Sawyer you are saying this allows me to do what Claude workflows do but with other models ?
@smehmood @niteshiftdev please do haha!
@morganlinton @sawyerhood 100% this
@sawyerhood Can i do this in your app @devindesktop
Siqi Chen#1537pro-tip: fable is really good at writing workflows, but very expensive.
let fable orchestrate a fleet of gpt-5.5 agents and get more bang for your buck!