Hi there. agentfab is distributed in the sense that it can run across different compute instances and has a control plane service managing nodes and agent instances. There's a blog post I wrote delving deeper into those capabilities with a more apt demo here: https://razvanmaftei.me/article?slug=agentfab-the-distribute...
As to your other question - agentfab uses a Conductor agent (baked into every fabric) that creates the taskgraph, which supports bounded review loops. The platform also supports pausing running tasks for user queries or change requests, which fires the decomposition step again. The end-to-end process can get a bit wordy to explain, my blog posts and the docs in the repo do a better job at that.
I'm 100% looking for collaborators on this, so if you're interested I'd love to discuss more.
I'm not exactly following as to who this is for - people are going to use email templates instead of writing Markdown emails, and agents can just as easily spit out HTML. Seems like your solution is in search of a problem.
I can't imagine they were getting a good return on it. And frankly, nothing tht came out of Sora was consequential in a positive way. The tech is cool, but only works if the content generation is heavily guardrailed and most of it ends up as content farming fodder anyway.
As to your other question - agentfab uses a Conductor agent (baked into every fabric) that creates the taskgraph, which supports bounded review loops. The platform also supports pausing running tasks for user queries or change requests, which fires the decomposition step again. The end-to-end process can get a bit wordy to explain, my blog posts and the docs in the repo do a better job at that.
I'm 100% looking for collaborators on this, so if you're interested I'd love to discuss more.