> ## Documentation Index
> Fetch the complete documentation index at: https://docs.youniverse.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Memory and playbooks

> How Agents learn between runs and how to reuse or share what they learned.

<Frame>
  <img src="https://chat.youniverse.ai/__l5e/assets-v1/593abab1-0bd7-40ef-9f88-0915eee8d91d/docs-agents-memory.jpg" alt="Notebook and a playbook card passed between robots" />
</Frame>

Agents get better every run. The **Memory** tab shows everything an Agent remembers.

## What it remembers

| Section | Contents |
| - | - |
| **What this Agent remembers** | Instructions you gave it, like "cap spending at \$200". Tell it in chat and it keeps it here. |
| **How it did this last time** | After a successful run it saves the approach that worked, so the next run is faster and cheaper. |
| **API notes** | Details it looked up about services, kept tidy so it doesn't search again. |

You can review and delete any memory.

## After every run

On the details page, **After every run** lets you add rules that run each time a real run finishes — for example "post a summary in Slack" or "update the tracker sheet". Previews and tests don't count.

<Frame>
  <img src="https://chat.youniverse.ai/__l5e/assets-v1/921e3baf-87cf-4795-8d9d-a991978fd5f1/docs-agents-create.jpg" alt="Glass robot typing a job" />
</Frame>

## Playbooks

A playbook packages what an Agent has learned so it can be reused.

* **Start a new Agent from this one** — clone the job and its learned approach.
* **Share playbook** — get a link others can import.
* **Paste a playbook link to import it** — then tap **Check** to preview before creating.

<Note>
  Passwords, vault items and memories are never copied into a playbook.
</Note>


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.