Your AI's memory needs a clear-out too

You have given your AI a memory. That could be a folder with notes or a file with what it needs to know about you. Perhaps it is even a fully fledged second brain in Obsidian. For the first few weeks, that works very well. You have to explain less and your agent knows where to look. And then one day you notice that it is slower at finding its way. Or that it brings up something that has not been correct for a long time.

In a vibe coding training course, I showed how my own agents work with such a memory. And I had to honestly admit that I ran into this myself. At a certain point, my vault started to become very complex indeed.

Copy the prompt at the bottom of this article. With it, you let your AI suggest an initial clear-out, without it throwing anything away just yet.

Why the memory of your agents grows by itself

There is a simple rule in the briefing for my agents. If they do something, they keep the memory up to date. That means that after its task, each agent writes a short log about what it has done. That is actually very handy. This way, an agent on my computer here knows what another agent on my computer at home has done.

And then it becomes a bit ironic. The better your agents do their work, the faster that memory grows. Every task leaves a trace. When I clicked open the folder with logs during the training, it had already become a very long list. The system that was supposed to keep everything simple became something to get lost in itself.

You might recognise this from your own notes, or from the shared drive at work. Nobody throws anything away, because everything was important once. And so searching becomes a little bit harder every month, for people and for AI.

Frequently asked questions

What do you mean by the memory of your AI?

Those are the files your AI reads to know your work. Think of a folder with Markdown notes or a vault in Obsidian. Your agents can also write in it themselves, for example a short log after every task.

Why should I throw anything away from the memory of my AI?

A memory that agents write in grows with every task. The bigger it gets, the harder it is for your agent to find what matters. A clear-out keeps what is important and lets the rest go.

Do I have to build a skill for that straight away?

No, start with a clear-out that you initiate yourself in a normal chat. Let your AI make suggestions first and decide for yourself what can go. If that goes well a few times, you ask it to turn it into a skill.

Where can I practise this hands-on?

In 'Claude AI Power User' and 'AI-agents' you set up a workspace and a memory for your agents yourself. You can find all dates at Workshops & Events.

Keep reading for free

Register free with your e-mail address and read on. Your account opens the free articles in the learning space; the rest is for subscribers, from Learn Plus up.

No comments yetRead the comments

Comments

Members comment here and read each other's comments. Register for free or sign in to join the conversation.

No comments yet. What do you take away?