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

# What's New in Liberator 2.5

> Release notes for the CloudQuant Data Liberator 2.5 release — Reporting tab, saved report templates, scheduled email delivery, admin activity audits, and cache setup during dataset creation

# What's new in Liberator 2.5

Liberator 2.5 adds a **Reporting** tab where you can run your own queries, save them as templates, and schedule those templates as reports delivered to an email address. Super Admins also get an **admin activity audit** of who changed users, entitlements, datasets, and other admin settings. You can turn on cache pre-generation while creating a dataset, and AI Assisted onboarding can reuse a connection you already have.

## Highlights

<CardGroup cols={2}>
  <Card title="Reporting tab" icon="file-lines" href="/administration/ai-based-reporting#reporting-tab">
    Run your own usage and entitlement queries, then save them as templates you can run again.
  </Card>

  <Card title="Scheduled email reports" icon="envelope" href="/administration/ai-based-reporting#schedule-a-template">
    Send a saved template on a daily, weekly, or monthly schedule to one or more email addresses.
  </Card>

  <Card title="Cache at dataset creation" icon="clock" href="/administration/cache-pre-generation">
    Turn on pre-generated cache in the Create Dataset wizard, not only when editing a dataset later.
  </Card>

  <Card title="Reuse an existing connection" icon="plug" href="/administration/ai-dataset-onboarding#reuse-an-existing-connection">
    AI Assisted onboarding can attach a new dataset to a connection you already tested and saved.
  </Card>

  <Card title="Admin activity audits" icon="clipboard-list" href="/administration/admin-activity-audits">
    See who changed users, entitlements, datasets, connections, and report schedules, including failed attempts.
  </Card>
</CardGroup>

## Reporting tab

Super Admins open **Reporting** in the admin portal. The page keeps the natural-language assistant and adds a list of saved templates.

* **Run a query** in the assistant, or use a quick-action report, and download the Excel result.
* **Save as Template** stores that query so you can run it again without retyping it.
* **Run** on a saved template downloads a fresh copy.
* Edit the template name, who it covers (everyone, one company, or one user), and its date window. Delete a template you no longer need.

A template needs a moving date window before it can be scheduled. Choose **Yesterday**, **Last 7 days**, **Last 30 days**, **Month to date**, or **Last month**. A fixed historical range stays available for one-off runs, and each scheduled run would otherwise resend the same dates.

See [AI-based reporting](/administration/ai-based-reporting).

## Scheduled email reports

From a saved template, open **Schedule** and choose when it runs:

* **Daily**, **weekly** (selected days), or **monthly** (a day of the month, the last day, or a weekday such as the first Monday)
* A time and timezone
* **Excel (.xlsx)** or **CSV**
* One or more recipient email addresses

Each run resolves the template's date window in the schedule's timezone, so "Yesterday" moves forward with the calendar. You can turn a schedule off without deleting it, or remove it entirely. The template list shows whether the last run succeeded.

Webhook delivery is also available when email is not the destination you want.

## Cache pre-generation during dataset creation

You can configure pre-generated cache in the **Create Dataset** wizard, under **Advanced options**, before you save the dataset. The same Parquet destination, rolling window, fixed date range, and full-dataset choices remain available when you edit a dataset later.

See [Cache pre-generation](/administration/cache-pre-generation).

## AI onboarding and OneDrive

* The assistant can **reuse an existing connection**. It skips creating a second connection when you name one that is already saved, then continues with dataset settings.
* **OneDrive** requires a folder prefix. A blank prefix no longer means the root of the signed-in account.

See [AI-driven dataset onboarding](/administration/ai-dataset-onboarding) and [SharePoint and OneDrive](/datasource-config/sharepoint-onedrive).

## Admin activity audits

Super Admins can review an activity audit of changes made in the admin portal: users, entitlements, datasets, connections, groups, webhooks, files, report templates, and report schedules. Each entry records when the change happened, who made it, what changed, and whether it succeeded. Failed attempts stay in the audit. Passwords and API tokens are omitted.

Ask from the **Reporting** tab, for example "who gave this user access" or "what changed last Tuesday." Changes made outside Liberator, including in the identity provider console, are not included.

See [Admin activity audits](/administration/admin-activity-audits).

## Administration

Usernames are stored in lowercase, so sign-in does not depend on the capitalization used when the account was created.

## Reliability

* Onboarding is more reliable at detecting date and time columns and formats.
* A query that disconnects mid-listing no longer leaves the next query on that dataset waiting.
* If one storage mount fails to attach, the other mounts still come up.

## Compatibility

There are no breaking changes to the client libraries or to the dataset query surface. Saved templates, schedules, and create-time cache settings are additions in the admin portal.