I’ve been exploring how multi-agent workflows could be useful for educational websites, especially when there are lots of similar tasks involved, such as researching topics, checking information, preparing outlines, reviewing content, and keeping older articles updated.
For example, I’m interested in experimenting with a workflow for a site like learnquranonlineusa.com, where the content is focused on online Quran learning and educational resources.
One idea would be to have separate agents handle research, content structure, fact-checking, and a final quality review, with a human checking the output before anything is published.
I’m curious how others here approach this. Do you find it better to give each agent a very specific responsibility, or have one stronger agent handle most of the workflow? Also, what are some good ways to prevent agents from repeating work or passing unnecessary context between tasks?