> For the complete documentation index, see [llms.txt](https://docs.ebrain.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.ebrain.ai/chatting-with-tim/memory.md).
# How Tim remembers things
Tim has a long-term memory. Across every conversation, and across web chat, WhatsApp, and your autonomous [Agents](/agents-autonomous/overview.md), he remembers useful facts about you and your business, your preferences, the decisions you've made, and the ways you like things done. The result: you stop re-explaining context every time you talk to him.
There are two sides to this:
1. **Things Tim remembers automatically** during normal conversations.
2. **A memory list you control** in Settings, where you can view, add, edit, and delete what Tim remembers.
{% hint style="success" %}
Memory is on by default. There's nothing to enable and no setup required, it works as soon as you start using Tim.
{% endhint %}
## What Tim remembers
Every remembered item has a **type**. There are four:
| Type | What it's for |
| -------------- | --------------------------------------------------------------------------- |
| **Fact** | Durable knowledge, like where your office is or when your fiscal year ends. |
| **Preference** | How you like things done, like "short, direct answers without jargon." |
| **Decision** | A past decision and its reasoning, so Tim doesn't later contradict it. |
| **Workflow** | A useful procedure or way of working that you taught him or he discovered. |
Tim aims to store concise, actionable information, not entire conversations and not obvious or throwaway details.
In the background, Tim also keeps lightweight notes of your daily activity and short summaries of past conversations. These help him stay oriented in time and recall the gist of older chats. They work entirely behind the scenes, so there's no screen for you to open.
## Everyday memory vs. always-on memory
Not everything Tim knows is loaded into his head at all times, that would be too much. There's an important difference:
* **Everyday memories** are stored and looked up when they're relevant. When you ask something, Tim searches his memory and pulls in what matters for that question.
* **Always-on (core) memory** is a small, curated set of your most important, frequently-used memories that Tim keeps in front of him every session, without needing to search.
You don't manage the always-on set by hand. eBrain runs a quiet, periodic housekeeping pass that **promotes** the memories you rely on most into the always-on set, **merges** duplicates so they don't pile up, and **retires** stale entries that have stopped being useful.
{% hint style="info" %}
A memory you just added isn't necessarily always-on right away, importance is earned through use. The practical takeaway: keep talking to Tim and saving useful things. The ones that matter keep getting easier for him to recall, and clutter gets cleaned up on its own.
{% endhint %}
## Personal vs. company memory
Memory comes in two scopes, shown as two tabs on the settings page:
{% hint style="info" %}
**Your memories** are facts and preferences Tim remembers about you. Only you can see these.
**Company memories** are knowledge Tim shares across everyone in your company or workspace.
{% endhint %}
When Tim saves something on his own, he defaults to **personal** ("Your memories"). He only files something under company memory when it's objectively true about the business no matter who said it, like a company-wide policy, a recurring meeting cadence, or a shared vendor.
{% hint style="warning" %}
Because a company memory is visible to your whole team, a wrong or overly personal entry there can mislead everyone. When in doubt, keep it personal, and don't put information meant only for you into a company memory.
{% endhint %}
If you belong to more than one workspace, personal and company memories are kept separate per workspace.
## Where to manage memory
Open **Settings**, and under the **Your Account** section in the left settings navigation, choose **Capabilities**. The page is titled **Capabilities** with the description "View and manage what Tim remembers about you and your company." There's no separate Memory item in the main sidebar, memory management lives here.
The page has two tabs, **Your memories** and **Company memories**, each listing its memories as cards. Every card shows the memory text, a type badge (Fact, Preference, Decision, or Workflow), and an "Updated" date.
## Asking Tim to remember something
You have two ways to add a memory.
{% hint style="info" %}
**Just tell him in chat.** Say things like:
* "Remember that I prefer morning meetings."
* "From now on, always CC my assistant on client emails."
* "Our office is in Amsterdam and our fiscal year ends in March."
Tim recognizes durable facts, preferences, and decisions and saves them himself. He'll also proactively save important things you share about your business, even without being asked.
{% endhint %}
You can watch memory at work in chat through Tim's activity indicators: **Searching memory** or **Checking memory** when he looks something up, **Recalling** when he pulls a specific item, **Remembering** when he saves something new, and **Logging activity** when he notes an event for context.
To add a memory yourself:
{% stepper %}
{% step %}
## Pick the right tab
On the **Capabilities** page, choose **Your memories** for something private to you, or **Company memories** for something the whole team should know. The scope is set by the tab you're on, so make sure it's the right one.
{% endstep %}
{% step %}
## Open the add dialog
Click **Add memory**. The dialog is titled "Add a personal memory" or "Add a company memory" depending on your tab.
{% endstep %}
{% step %}
## Write the memory and choose a type
Type the fact or preference in the **Memory** box, then pick a **Type** (Fact, Preference, Decision, or Workflow).
{% endstep %}
{% step %}
## Save
Click **Save memory**. You'll see a "Memory saved" confirmation.
{% endstep %}
{% endstepper %}
## Viewing, editing, and deleting memories
All of this happens on the **Capabilities** page.
How do I edit a memory?
Click the edit (pencil) icon on the memory. The "Edit memory" dialog opens, change the text and click **Save changes**. You'll see "Memory updated."
How do I make Tim forget something?
Click the delete (trash) icon on the memory. A confirmation dialog warns that "Tim will forget this permanently. This cannot be undone." Click **Delete memory** to confirm, and you'll see "Memory deleted."
The Delete control on the Capabilities page is the reliable, permanent way to remove a memory. If you'd rather ask Tim to forget something in chat, still confirm the removal here so it's gone for good.
{% hint style="warning" %}
Deletion is permanent and cannot be undone.
You can always manage your own personal memories. **Company memories can only be removed by an admin or owner**, regular members can view and add them but can't delete them.
{% endhint %}
## Where your memories get used
* **Everywhere you talk to Tim**, both the [Chats page](/chatting-with-tim/overview.md) and [WhatsApp](/ebrain-on-whatsapp/getting-started.md) draw on the same memory.
* **Across separate conversations**, context carries over even into a brand-new chat.
* **In specialist work and autonomous Agents**, your always-on memory travels with Tim into [specialist tasks](/chatting-with-tim/thinking-and-specialists.md) and scheduled [Agent runs](/agents-autonomous/scheduling-and-runs.md), so your standing preferences and key facts apply there too.
* **Proactively**, because core memories are always in context, Tim's nudges and drafts reflect what he knows about you, like your preferred tone.
## Good habits
* Review the Capabilities page now and then. Tim decides what's worth keeping, so it's worth editing or deleting anything that's inaccurate.
* Keep sensitive, personal-to-you information in **Your memories**, not company memory.
* Double-check the active tab before adding, it decides who can see the memory.
## Related pages
---
# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.
## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.
Perform an HTTP GET request on the current page URL with the `ask` query parameter:
```
GET https://docs.ebrain.ai/chatting-with-tim/memory.md?ask=
```
The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.