> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs-parschat.parstechai.com/user-manual/knowledge/add-sources/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs-parschat.parstechai.com/_mcp/server.

# Add data sources

> Build your robot’s knowledge from files, web pages and online stores.

Your robot answers from the sources you give it. Open your robot and go to **Train Bot** (in the
training menu). The page has three tabs: **Add File**, **Add Link** and **Add ecommerce website**.

## Add files

#### Choose your files

On the **Add File** tab, upload one or more files. Accepted formats are **PDF**, **DOC**,
**DOCX** and **TXT**, up to **10 MB per file**. The same file can’t be added twice.

![Add File](/_fern-img/14f374884c74a2c8a334837ed15aa6d34bb46ee9261a821235624cef941c8de0.webp)

#### Train

Click **Train Bot**. When the upload finishes you are taken to
[Data Collection](/user-manual/knowledge/manage-sources), where you can follow training.

## Add web pages

#### Choose what to read

On the **Add Link** tab, choose whether to read a single **Page** or crawl the **whole
website**.

![Add Link: a single page or a whole website](/_fern-img/47a9ee36367b9b356082cb232bc16e2fed5cdbed6689596a6f6d1e88b00fc760.webp)

#### Enter the address

In **Add your link**, enter the address. Only **HTTPS** addresses are accepted; `https://` is
added for you if you leave it out.

#### Adjust the crawl (optional)

Under **Advance Settings** you can set:

| Setting                                   | What it does                                                 |
| ----------------------------------------- | ------------------------------------------------------------ |
| **Number of links that can be extracted** | How many pages to collect, up to 50                          |
| **Departure time**                        | How long the crawl may run, in minutes (up to 50)            |
| **Ignore the sitemap**                    | Crawl by following links instead of using the site’s sitemap |
| **Extract sub-links**                     | Also collect pages linked from the ones found                |

#### Pick pages and train

Click **Add Link**, choose the pages you want (up to **50**), then click **Train Bot**.

## Add an online store

The **Add ecommerce website** tab reads products and store content directly from your shop:

![Add ecommerce website](/_fern-img/6bab4ed354e6a754f32a93d5c5c97da3714d512041930f8318a721422569f1aa.webp)

| Platform                    | Guide                                                                  |
| --------------------------- | ---------------------------------------------------------------------- |
| **WordPress & WooCommerce** | [Train from a WordPress store](/user-manual/knowledge/wordpress-store) |
| **Mixin Website Builder**   | [Train from a Mixin store](/user-manual/knowledge/mixin-store)         |
| **Digify**                  | Connect it from the same tab                                           |
| **API**                     | Connect a custom store API from the same tab                           |

PrestaShop and Portal are listed as coming soon.

## Add questions and answers by hand

To teach the robot a specific answer without writing a document, open **Q\&A** in the training
menu, click **Add**, and enter a question and its answer. Click **Apply** to save it.

> **Tip**
>
> Test what the robot has learned from **Chatbot Test** in the training menu before you put it in
> front of customers.

> Build your robot’s knowledge from files, web pages and online stores.