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

# Get started with Auto Capture

> Your first Auto Capture video in six steps: check your workspace has access, save a demo login if the page needs a sign-in, ask the Tella Agent or a connected AI app for a take, then review the handoff and edit the video.

> [!WARNING]
>
> Auto Capture is in a **closed beta** and isn't available to every workspace yet. We'd love your feedback: email [support@tella.com](mailto:support@tella.com).

To make your first Auto Capture video, check your workspace has Auto Capture, save a demo login if the page is behind a sign-in, then ask the Tella Agent in the editor, or an AI app connected to Tella, to record the page. A few minutes later you get a finished screen recording with an animated cursor, ready to edit and share. Nobody has to sit down and record.

[https://www.tella.tv/video/vid\_cmvflam0t001g02oy1tvqc33l/embed?b=0\&title=0\&a=1\&loop=0\&t=0\&muted=0\&wt=0](https://www.tella.tv/video/vid_cmvflam0t001g02oy1tvqc33l/embed?b=0\&title=0\&a=1\&loop=0\&t=0\&muted=0\&wt=0)

Auto Capture filmed the video above from this prompt and the app's sign-in page URL, with nothing else:

```text wrap theme={"system"}
Record a complete product video of a plumber's morning in Tallyfern, from the sign-in screen: sign in with email, password and the authenticator code, glance at what's overdue on the dashboard, then create an invoice for a customer with a few line items and a discount, preview it as the customer sees it, send it, and finish on the sent invoice. Film the whole journey on camera, page to page, rather than jumping straight to the end.
```

The app is a fictional demo product, so its names and figures are made up. It typed the password and the authenticator code itself. There are more prompts and their videos in [Auto Capture example prompts](/docs/help/integrations/auto-capture-example-prompts.md).

## Step 1: Check your workspace has Auto Capture

1. Click your **workspace name** in the top-left corner and choose **Settings**.
2. Look for **Credentials** under **Other** in the settings sidebar.

If **Credentials** is there, your workspace has Auto Capture. If it isn't, Auto Capture isn't available for your workspace yet: email [support@tella.com](mailto:support@tella.com) to ask about the beta.

Each take uses your workspace's [credits](/docs/help/billing/ai-credits.md), the same balance the Tella Agent spends. See [Where do I see my credit balance?](/docs/help/billing/ai-credits.md#where-do-i-see-my-credit-balance)

## Step 2: Pick a page and a journey to film

Choose one flow you'd otherwise record by hand, and the full URL of the page it starts on. Good first takes:

* **A public page:** your pricing page, a landing page or a docs page. No login is needed, so it's the quickest way to see Auto Capture work.
* **A feature in your product:** something a customer does in a few steps, such as creating a project, inviting a teammate or exporting a report.

One take films one journey. It can cross several pages, but keep it to a single task. For a longer video, ask for one take per task and add them to the same video.

## Step 3: Save a login if the page needs a sign-in

Skip this step for public pages. For pages behind a sign-in:

1. In **Settings** → **Credentials**, click **Add credential**.
2. Choose **Enter manually**, or pick a login from 1Password.
3. Give it a **Name** such as "Demo account", set **Sign-in page** to the site where the password is typed, and check **Use on** covers the site your take starts on.
4. Enter the **Email** and **Password**, then click **Add credential**.

Use a demo account with realistic sample data, not your own account or one with customer data: everything the account can see may end up in the video. If your product asks for a code after the password, save the login from 1Password with its one-time password, and Auto Capture fills the code in itself. For every option, see [Let Auto Capture sign in to your product](/docs/help/integrations/auto-capture-credentials.md).

> [!WARNING]
>
> Never paste a password, code or sign-in link into a chat. Auto Capture only signs in with logins saved under **Settings** → **Credentials**.

## Step 4: Ask for your first take

**In the Tella editor, with the Tella Agent:**

1. Open a video in the editor. A new, empty video is fine.
2. If the sidebar on the right is hidden, press **Shift + S** to open the [Tella Agent](/docs/help/editing/edit-with-the-tella-agent.md).
3. Type your request with the full URL, and press **Enter**.

**In Claude, ChatGPT or another AI app:**

1. Connect the app to Tella in **Settings** → **Connections**, under **AI apps**. See [Connect an AI assistant to Tella](/docs/help/integrations/mcp-server.md#connect-an-ai-assistant-to-tella).
2. In a chat with the Tella connector turned on, type your request with the full URL.

Write the request the way you'd brief a colleague who's never seen your product:

```text wrap theme={"system"}
Use Tella Auto Capture to make a product demo of https://app.example.com
for people trying us for the first time: create a new project, invite a
teammate to it, and finish on the project page with them listed.
Sign in with the "Demo account" credential. Film the whole journey on
camera, page to page, rather than jumping straight to the end.
```

* **Say who it's for**, such as "first-time visitors" or "admins setting up SSO".
* **Describe the journey and where it ends**, not the clicks. Auto Capture finds the buttons itself.
* **Ask for the whole journey on camera** if viewers need to see how to get there. Otherwise Auto Capture may get to the feature off camera and start filming on the final screen.
* **Say what not to do** if it matters: "don't send anything", "don't change settings".
* **Leave out a length** unless you need one. Without one, the take is as long as the journey.

For more ideas, see [Auto Capture example prompts](/docs/help/integrations/auto-capture-example-prompts.md).

## Step 5: Wait for the take, and answer it if it asks

A take usually takes a few minutes. It explores the page, writes a script, checks the script runs, then films it in a clean browser. Keep working while it runs; the Tella Agent or your AI app checks on it for you. In some AI apps, you may need to ask "Is the Tella capture done?"

A take can pause or stop to ask you something:

* **A sign-in code:** you get a link to a Tella page. Type the code your product sent you there, and the take carries on. See [Enter a sign-in code for Auto Capture](/docs/help/integrations/auto-capture.md#enter-a-sign-in-code-for-auto-capture).
* **A missing login:** you get a link to **Settings** → **Credentials**. Add the login (step 3), then ask for the take again.
* **A question:** such as which of two projects to use. Answer it in the chat, and a new take starts with your answer.

## Step 6: Review the take and make it a video

When the take is delivered, you get a **handoff note**: what the video shows, what Auto Capture did off camera, and anywhere it departed from your request. Read it before you use the take.

* **In the editor,** the Tella Agent places the take in your video and applies the suggested cuts it agrees with.
* **In an AI app,** ask it to "put the take in a new Tella video". The take isn't in your library until it's in a video.

Then open the video in Tella and edit it like any screen recording: trim it, add [zooms](/docs/help/editing/add-a-zoom.md), [restyle the cursor](/docs/help/editing/customize-animated-cursor.md), speed it up or record a voiceover over it.

## Change your first take

If the take is almost right, ask for a revision and say what to change: "Revise that take so it also opens the invite email preview and ends there." A revision starts from what the first take learned, so it's quicker and uses fewer credits than starting over. See [Change or redo an Auto Capture take](/docs/help/integrations/auto-capture.md#change-or-redo-an-auto-capture-take).

## What to do after your first take

* Make a take for each step of a longer tour, and add them to one video in order.
* Record a video for each of your help articles: see [Make videos for your help articles with Auto Capture](/docs/help/integrations/auto-capture-help-center-videos.md).
* If your site shows a bot check to Auto Capture, let it through: see [Let Auto Capture past your firewall or bot protection](/docs/help/integrations/auto-capture-site-access.md).


## Related topics

- [Auto Capture example prompts](/docs/help/integrations/auto-capture-example-prompts.md)
- [Record a product demo with Auto Capture](/docs/help/integrations/auto-capture.md)
- [Is Auto Capture safe to give my sign-in details?](/docs/help/integrations/auto-capture-security.md)
- [Let Auto Capture past your firewall or bot protection](/docs/help/integrations/auto-capture-site-access.md)
- [Let Auto Capture sign in to your product](/docs/help/integrations/auto-capture-credentials.md)


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