fixed a bug that inner fluxes inside flatMap callbacks were not being awaited until completion #monoflux

#monoflux was not running all flatMap callbacks as soon as source Flux emitted a value, but sometimes waiting until entire source Flux completed – now solved

#monoflux flatMap just got an optional concurrency parameter, with which you can set the maximum number of concurrently running callbacks

Time-non-critical agents are super slow because you use flex mode in OpenAI to save costs. Now with gpt-5 they became even slower (and cheaper, so it's a good thing!). So you *need* to parallelize your agents. This is what I've been doing last night. My library #monoflux's flatMap method on a Flux (a Flux is an asyncronous array) can now handle () => Flux callbacks. It is now my number 1 engine to run *tons* of agents in parallel.

Made .flatMap method process its callbacks in parallel #monoflux