Keep control

Understand your Mind’s settings

Your Mind is the Mensara workspace holding your team’s saved information and work. Its settings control different parts of that workspace: who may use information, when a change needs review, and which models or connected services can be used. This page helps you find the relevant control and understand why two Minds, or two people in the same Mind, can get different results.

Contents

Find Settings in your Portal

The Portal is your Mind’s web interface. Open Settings there. Under Workspace, Members takes an owner or admin to the page headed Team access. The desktop settings menu also calls that destination Team access.

The controls you see depend on your role, delegated permissions and the features available in that Mind. An admin may be able to inspect access without being allowed to change it. If a control is absent or read-only, ask the owner to check the relevant permission rather than assuming it is switched off.

Why changes wait for review

A proposal is a suggested change to a saved record. Before saving it, the Mind checks the rules for the area, the access of the person proposing it, the kind of change and the required checks. These rules are part of your Mind’s settings. They can allow an eligible correction to save immediately or require someone with approval authority to review it.

Passing one check is not a promise that a proposal will save. A protected kind of change, an older source version or an unavailable required check can still leave it waiting or prevent it from being accepted. Read the actual result. Applied means the change was saved; held means it is waiting. Review changes and check the saved result explains the next step.

The Portal does not provide one general editor for every area’s saving rules. The owner can use the team access controls described below; more detailed policy changes need the owner’s supported configuration route. Correcting a record walks through an ordinary correction without changing those settings.

The same edit can have a different outcome

The Mind checks more than whether a sentence looks correct.

The Mind checks the area rules, the proposer’s access, the kind of change, the current source and required checks. An eligible proposal can be saved immediately. A proposal that needs a decision is held for an authorised reviewer. A refused proposal has not saved. Read the result; the separate Automatic approval setting does not override these boundaries.

Team access presets

On Team access, the owner can inspect Team access preset and choose Switch to open team or Switch to locked down. A Mind can also have a custom arrangement.

Open team provides broader default access to shared areas and delegates some management to admins. It can make eligible owner or admin edits save after their checks. It does not mean every record is visible or every change saves automatically.

Locked down removes the broad Everyone default and the admin delegations controlled by that preset. It does not erase access already assigned directly or through a group. Check the person’s actual access as well as the preset. Roles, profiles and effective access explains the difference.

The owner can also decide whether admins may approve ordinary changes in areas they can read. Managing access, changing roles, running loops and other administrative jobs have separate delegations. One admin permission does not grant all the others.

Automatic approval

Automatic approval is a separate setting for eligible review cards that you are already allowed to approve. It can act on your behalf after a card exists. It does not give you more authority and is not a switch that makes every correction save immediately.

You can choose the eligible areas and whether the rule covers anyone’s proposals or only ones written by you and your agents. The Mind still checks your current authority, the source and the required checks before acting.

Some changes are excluded: special memory records wherever the Mind’s dedicated memory system is configured, policy and app code, deletions, and requests that explicitly require a human decision. This exclusion covers records managed by that dedicated memory system, including 22-memory; it does not mean every ordinary note saved in your Mind.

You can turn your own rule off. The owner can inspect who has it enabled and switch another person off, but cannot widen that person’s rule for them. Undo submits a checked reversal; it can need review itself or be refused if the source has changed. Open the record and inspect the result rather than assuming Undo restored it.

Models and spending

Your personal chat model, its provider connection and an installed app’s model access are different choices. Connecting the AI tool you already use does not by itself configure a model for an installed agent.

Settings shows the personal model connection, with Change model and Change provider routes. Owner-only App model access controls which connected routes and models apps may use; the app’s own spending budget still applies. The Usage display shows spending, not a promise that its visual scale is an enforced budget. An app’s model budget is checked against the previous seven days of usage before a call starts, so the final allowed call can cross a limit.

Before allowing an app to run, check its agreed model route, permissions, schedule and rolling seven-day model budget. Build and check an agent explains those prerequisites. Changing a model connection does not change who may approve a document.

Connections and backup

Connections manages external services and keys apps can use. A service connection is different from permission to read a saved area or use a model. Connections and capture explains bringing information in; Apps, sensors and connectors helps you choose a supported extension.

Backup and Recovery kit concern keeping and recovering your Mind. Follow the backup and recovery guide when deciding where the copy lives and what you need to restore it.

Check the effect of a change

Change one setting for a specific reason, then check the behaviour it should affect. For access, use the teammate’s own account and a harmless allowed task. For a proposed edit, inspect whether it was saved or held and read the source. For an installed agent, inspect its actual output and spending.

Keep these checks separate. Being able to open a page does not prove that a scheduled agent ran, and a saved correction does not prove that a teammate has the intended access.

Related

Practical guides for people and agents

Start with one useful job

Join the waitlist

Plain text for AI agents. Same content as the human page.

## For agents

- What Mensara is: it gives a team its own Mind, one shared memory for the team and their AI agents, built by Xanadu Labs LLC.
- Status: signup is not open. A person joins the waitlist at https://app.mensara.ai/#waitlist. Agents cannot create an account or sign up.
- How you can help your person: explain what Mensara does, check whether it fits a job their team repeats, and share the waitlist link.
- Agent references: https://app.mensara.ai/llms.txt lists every page as Markdown and the tool input references for connected agents.

---

# Understand your Mind’s settings

How Mensara works · Keep control · [How it works](https://app.mensara.ai/how-it-works)

Your Mind is the Mensara workspace holding your team’s saved information and work. Its settings control different parts of that workspace: who may use information, when a change needs review, and which models or connected services can be used. This page helps you find the relevant control and understand why two Minds, or two people in the same Mind, can get different results.

[Page as Markdown](https://app.mensara.ai/how-it-works/mind-settings.md)

## Find Settings in your Portal

The **Portal** is your Mind’s web interface. Open **Settings** there. Under **Workspace**, **Members** takes an owner or admin to the page headed **Team access**. The desktop settings menu also calls that destination Team access.

The controls you see depend on your role, delegated permissions and the features available in that Mind. An admin may be able to inspect access without being allowed to change it. If a control is absent or read-only, ask the owner to check the relevant permission rather than assuming it is switched off.

## Why changes wait for review

A **proposal** is a suggested change to a saved record. Before saving it, the Mind checks the rules for the area, the access of the person proposing it, the kind of change and the required checks. These rules are part of your Mind’s settings. They can allow an eligible correction to save immediately or require someone with approval authority to review it.

Passing one check is not a promise that a proposal will save. A protected kind of change, an older source version or an unavailable required check can still leave it waiting or prevent it from being accepted. Read the actual result. **Applied** means the change was saved; **held** means it is waiting. [Review changes and check the saved result](https://app.mensara.ai/how-it-works/changes-approvals-receipts#review-one-proposal) explains the next step.

The Portal does not provide one general editor for every area’s saving rules. The owner can use the team access controls described below; more detailed policy changes need the owner’s supported configuration route. [Correcting a record](https://app.mensara.ai/how-it-works/suggest-a-correction) walks through an ordinary correction without changing those settings.

### The same edit can have a different outcome

The Mind checks more than whether a sentence looks correct.

The Mind checks the area rules, the proposer’s access, the kind of change, the current source and required checks. An eligible proposal can be saved immediately. A proposal that needs a decision is held for an authorised reviewer. A refused proposal has not saved. Read the result; the separate Automatic approval setting does not override these boundaries.

## Team access presets

On **Team access**, the owner can inspect **Team access preset** and choose **Switch to open team** or **Switch to locked down**. A Mind can also have a custom arrangement.

Open team provides broader default access to shared areas and delegates some management to admins. It can make eligible owner or admin edits save after their checks. It does not mean every record is visible or every change saves automatically.

Locked down removes the broad Everyone default and the admin delegations controlled by that preset. It does not erase access already assigned directly or through a group. Check the person’s actual access as well as the preset. [Roles, profiles and effective access](https://app.mensara.ai/how-it-works/right-access#roles-and-profiles) explains the difference.

The owner can also decide whether admins may approve ordinary changes in areas they can read. Managing access, changing roles, running loops and other administrative jobs have separate delegations. One admin permission does not grant all the others.

## Automatic approval

**Automatic approval** is a separate setting for eligible review cards that you are already allowed to approve. It can act on your behalf after a card exists. It does not give you more authority and is not a switch that makes every correction save immediately.

You can choose the eligible areas and whether the rule covers anyone’s proposals or only ones written by you and your agents. The Mind still checks your current authority, the source and the required checks before acting.

Some changes are excluded: special memory records wherever the Mind’s dedicated memory system is configured, policy and app code, deletions, and requests that explicitly require a human decision. This exclusion covers records managed by that dedicated memory system, including `22-memory`; it does not mean every ordinary note saved in your Mind.

You can turn your own rule off. The owner can inspect who has it enabled and switch another person off, but cannot widen that person’s rule for them. **Undo** submits a checked reversal; it can need review itself or be refused if the source has changed. Open the record and inspect the result rather than assuming Undo restored it.

## Models and spending

Your personal chat model, its provider connection and an installed app’s model access are different choices. Connecting [the AI tool you already use](https://app.mensara.ai/how-it-works/portal-mcp-cli) does not by itself configure a model for [an installed agent](https://app.mensara.ai/how-it-works/build-an-agent).

Settings shows the personal model connection, with **Change model** and **Change provider** routes. Owner-only **App model access** controls which connected routes and models apps may use; the app’s own spending budget still applies. The Usage display shows spending, not a promise that its visual scale is an enforced budget. An app’s model budget is checked against the previous seven days of usage before a call starts, so the final allowed call can cross a limit.

Before allowing an app to run, check its agreed model route, permissions, schedule and rolling seven-day model budget. [Build and check an agent](https://app.mensara.ai/how-it-works/build-an-agent#before-you-start) explains those prerequisites. Changing a model connection does not change who may approve a document.

## Connections and backup

**Connections** manages external services and keys apps can use. A service connection is different from permission to read a saved area or use a model. [Connections and capture](https://app.mensara.ai/how-it-works/connections-and-capture) explains bringing information in; [Apps, sensors and connectors](https://app.mensara.ai/how-it-works/apps-sensors-connectors) helps you choose a supported extension.

**Backup** and **Recovery kit** concern keeping and recovering your Mind. Follow [the backup and recovery guide](https://app.mensara.ai/how-it-works/backup-restore-ownership) when deciding where the copy lives and what you need to restore it.

## Check the effect of a change

Change one setting for a specific reason, then check the behaviour it should affect. For access, use the teammate’s own account and a harmless allowed task. For a proposed edit, inspect whether it was saved or held and read the source. For an installed agent, inspect its actual output and spending.

Keep these checks separate. Being able to open a page does not prove that a scheduled agent ran, and a saved correction does not prove that a teammate has the intended access.

## Related

- [Choose access for a teammate](https://app.mensara.ai/how-it-works/right-access)
- [How zones and permissions work](https://app.mensara.ai/how-it-works/zones-and-permissions)
- [Review changes and check what was saved](https://app.mensara.ai/how-it-works/changes-approvals-receipts)
- [Set up your Mensara Mind and get a first useful answer](https://app.mensara.ai/how-it-works/setup-and-ongoing-care)

Previous page: [Review changes and check what was saved](https://app.mensara.ai/how-it-works/changes-approvals-receipts)

Next page: [Back up, recover and export your Mind](https://app.mensara.ai/how-it-works/backup-restore-ownership)

A backup is a second copy kept away from the server running your Mind.

[Join the waitlist](https://app.mensara.ai/#waitlist)