Heavy agree. For high level architecture work, refactors, or any other high int problems, I use Fable and Astra. Whichever one starts the work, the other reviews the work / makes suggestions.

There's a trend on twitter going around telling you to paste the following into your agent, and it's the kind of work I go straight to fable+astra for:

    Find all of the implicit state machines in my app. Make recommendations for changes that will elevate all three: the user experience (UX), developer experience (DX), and agent experience (AX) always, without breaking anything.

    Look at the code pathways for anything you find. Look through the git history for each one. Understand why code was written the way it was. Revisit each hypothesis and give me a full plan again after you're done.

Absolutely. Opus really sucks at broad context. I guess I could go back and forth with Opus editing out all its little mistakes or oversights in high-level planning, but I find it much easier to just use Fable for it instead. Basically, I wouldn't use Opus for audit work either. It seems too likely to come back with imaginary issues or hallucinate that everything is great when there's a glaring security flaw. Fable, OTOH, gives me neat lists of details to probe. Fable recommendations are usually reasonable, but sometimes they're prone to over-engineering.