# Agent memory - simple solution — retained record

> Retained reddit record from reddit:ClaudeCode.

Canonical: https://highsignal.app/data/records/39107f28fe057af1

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:ClaudeCode)  published 2026-10-21 15:52 ingested 2026-10-22 02:42

[original source ↗](https://www.reddit.com/r/ClaudeCode/comments/1xbqwdp/agent_memory_simple_solution/)

## retained text



```
I found a surprisingly simple way to give your AI agent a memory, and it doesn’t cost anything.

The problem is pretty familiar. You have a great session with Calude, you fix a tricky bug, and then you start a new session and have to explain the whole project all over again: how it’s structured, where the files live, what you’ve already tried. That’s because every new session starts with a blank memory.

There are plenty of solutions out there, but most of them mean sending your data to someone else’s server or paying for a plan. I came across a much simpler idea that actually works.

Here’s the setup. Create a local Obsidian vault on your hard drive, for example `c:\path\agents-wiki`. Then add two instructions to your [`agents.md`](http://agents.md) file:

1. Before asking the user any question, check the `agents-wiki` first to see if the answer is already there.
2. During or after a task, if you find something that would be useful for future sessions, write it to the `agents-wiki`.

That’s it. The agent gets a shared notebook that persists between sessions, it lives on your own machine, and Obsidian makes it easy to read and edit the notes yourself when you want to.
```



document fetched 2026-10-22 00:43 · canonical https://reddit.com/r/ClaudeCode/comments/1xbqwdp/agent_memory_simple_solution
