Testing Teamily AI 2.0's Loop multi-agent writing workflow
The author ran an AI news post through three agents plus himself: finding the story, checking the original source and drafting, with the agents correcting each other but the final call staying with the human.
I tested Loop in @Teamily_AI 2.0 on my AI news post. 3 agents and me. News Finder finds the story, Researcher checks it on the original page, Writer drafts the LinkedIn post, then it comes to me. one @ on News Finder and it ran on its own. it picked OpenAI's dots launch. the Researcher checked it on OpenAI's own page and left a note for the Writer: "do NOT say unlimited". honestly the Writer still wrote "skips usage limits". so I told it to use OpenAI's exact wording and cut a couplet. it fixed both and I approved it. one chatbot doesn't push back on itself like that. what stood out wasn't just the automation. it's a team workflow I can run again. the agents check each other, share the context (I didn't paste anything between them), and the final call still comes back to me. try it yourself, link in the reply.