I am trying to improve my approach to Research Agents and could use some practical advice.
I am particularly interested in testing and prompt design. I would like to compare approaches that people have actually used rather than just theoretical suggestions.
Which approach would you recommend...
