# Org settings

> Every security and data setting in Settings, where it is, its default, and what it does.

Only org admins see Settings.

## Settings

| Setting | Where | Default | What it does |
| --- | --- | --- | --- |
| **Turn on connections that change other systems automatically** | **Settings → Systems → Connections that change other systems** | **On (auto-enable)** | While on, a new write operation is granted as soon as it's registered. Off, each new write waits until a person approves it once. See [Security and approval](https://docs.mindset.ai/docs/security-model) |
| **What happens to personal data in transit** | **Settings → Governance → Personal data** | Off | Replaces personal data with placeholders before it reaches the model provider. See [Audit and personal data](https://docs.mindset.ai/docs/audit-and-personal-data) |
| **Audit retention (months)** | **Settings → Governance → Audit & data residency** | 6 | How long the audit record keeps identifying details |
| Data residency region | **Settings → Governance → Audit & data residency** | Set when the org is created | Shown only. It can't be changed |
| **Telemetry export (OpenTelemetry)** | **Settings → Governance** | Not set up | Sends traces, metrics, agent behavior and costs to your own collector. See [See what happened](https://docs.mindset.ai/docs/check-what-happened#alert-from-your-own-tooling) |
| **Conversation retention** | **Settings → Governance** | Kept indefinitely, except test runs (30 days) | When conversations expire after inactivity |
| Erase conversations | **Settings → Governance → Danger zone** | Nothing erased | Erases every end-user conversation and unattended run in the org, after a short grace period in which you can cancel |
| **Run the built-in agents on our own key** | **Settings → Governance → Keys for the built-in agents** | Off (platform-provided keys) | Moves Orca and the builders onto your org's own model key. See [Use your agents from AI clients](https://docs.mindset.ai/docs/use-your-agents-from-claude-and-other-ai-clients) |
| Stop a connection | The connection's **Overview** tab | Enabled | **Disable** refuses every call to the connection before anything leaves |

## Audit retention

| Value | Rule |
| --- | --- |
| Minimum | 6 months |
| Maximum | 120 months |
| When records pass it | Their identifying details are removed. The records themselves are never deleted |

## Telemetry export

Enter your **OTLP endpoint**, and an **Auth header name** and **Auth token** if your collector needs them.

| Switch | Default |
| --- | --- |
| **Send traces** | On |
| **Send metrics** | On |
| **Send agent behavior (script phases and gate outcomes)** | On |
| **Include end-user identifiers** | On |
| **Send costs** | Off. Sent only while **Send traces** is on |

## Conversation retention

Set separately for end-user conversations and unattended runs, rehearsals, authoring conversations and test runs. Each is **Kept indefinitely** or expires after a number of days without activity.

Per-call approval of one action isn't a setting. A script author builds it as an `$ask` step. See [Security and approval](https://docs.mindset.ai/docs/security-model).