This page was translated from Dutch with AI.
Fewer agents, more skills
In a workshop about agents, a question came up that took me a long time to figure out myself. One agent can have multiple skills, but why not just build multiple agents that each do one thing well? Surely that seems more organised.
That is how we did it too, not so long ago. An agent for every task. One for translations, one for newsletters, one for quotes, one for reports. Each with its own instructions and its own conversations. It felt organised. Until it was not anymore.
Where things went wrong
Eventually I had so many that I could no longer count them. And in each agent there were different chat sessions. I no longer knew who I had asked what. Some were on my laptop, others on a computer at home, and I had to search all over again every time to find which one was where.
Worse still: they all had to know the same things. Who I am, who I write for, how my texts sound. That was in every agent again, slightly different each time, and if I changed something, I had to change it everywhere. That was not just my problem. Everyone who worked like this ran into it, and the AI companies have adapted their tools to it in the meantime.
Keep reading for free
Register with your e-mail address to read the whole article. It is free, and it opens every article and programme in the learning space.
Check your inbox
We've sent you a sign-in link. It works for 30 minutes. Nothing there? Check your spam folder too.