You can understand the essence of the process by imagining that every action in the browser (clicking, pressing a button, entering login/password, etc.) can be assigned to a robot. It is enough to create a "script" - a sequence of actions that the "robot" will perform.

For example: we need a robot to perform some action on all FB accounts (and there can be any number of them), for example, confirming the binding of a card. If this were done manually, the task would take several hours - you have to log in to each profile, perform a number of actions, and then log out. When there are a lot of accounts, it turns into a big waste of time. With this functionality you can create a simple script that will perform all the necessary actions automatically 😉

The scenarios are based on possible everyday actions that usually take up a lot of time during work. A sequence of "actions" is created and automatically executed in the browser profile.

​​Elements ⬇️

Deleting any action (For example: Close tab) is performed after double-clicking on a block or selecting by pressing Shift to select several blocks at once, then press Del on the keyboard to delete.

Connecting the first block to the green circle "Start"
This makes it clear to the script where the starting point of the process starts.
All actions should be connected with 🧵 Thread between them, so that the Designer can understand the sequence of actions.
​Threads can be canceled by clicking ❌ or by connecting 2 in parallel.
​Red thread - gives the possibility to choose. Red thread is a safety net, it sets another action in case the Green thread didn't work or there was a problem on the site and the button the script should have clicked was not displayed. Then for example: the script will follow the Red Thread, which will lead it to some block.

Some action blocks use system variables containing browser profile parameters as input data. This is necessary for mass logins to major social networks or for creating scripts that interact not only within the browser, but also between profiles. System variables cannot be modified and cannot be used as output.

Navigation is all about interacting with web pages: opening or refreshing a page, navigating to a URL, switching between tabs, and other actions. Anything that is normally done through the address bar can be automated.

For example, if the first action in the browser is to open a specific web page, just specify the URL and the robot will open it automatically.

A handy tool for working in a team, it allows the script creator, when sharing the script with all the users in the team, to have each of them specify their own action in the element, instead of the one specified by the script creator.

Mouse is all the actions aimed at reproducing the cursor movements as accurately as possible. Clicking on a certain button, scrolling to the desired element on a web page - everything is easy to perform, you just need to set the necessary parameters.

For example, after opening a web page, you need to click on a certain field or button to enter data. It is enough to specify coordinates or a page element to be clicked by the bot.

There are 3 types of Selector ⬇️

XPath - allows you to search by any attribute of an element. Developers often add their attributes to various tags, which can be used, for example, to search the DOM tree, select author blog posts, or search tables. XPath is effective where other search methods fail.
​
Tables are found on many web pages, and there are often many of them on a single page, each of which can contain hundreds of rows. How do I make a selector for a link in an "In Concordance" cell? Classes can be at the top levels, tags are the same, and there are no attributes at all. XPath does a great job with this because of its features and full-text search capability. For example:

Such a selector is easy to read and easy to modify if necessary.
​

In addition, XPath demonstrates flexibility. If, in the example above, the set of records would contain 'Procurement', you could add the procurement number as an additional condition, e.g. as follows:

//*[contains(.,'Transactions')]//tr[contains(.,'Procurement 4')]//td[contains(.,'Under approval')]//a

You can get XPath on a page in the browser as follows: press CTRL+Shift+C (or on Mac - CMD+Shift+C), then select the desired element on the page and click on it. After that, the code of the desired element will be highlighted, from which you can copy the path in XPath format

​CSS is the selection of an element through its value in a stylesheet. A more convenient and frequently used selector. Read More

​Text - a simplified version of those listed above as it simply selects an item via text on the specified page

2. Mouse movement - works in the same way as coordinates.

3. Scrolling - simulates scrolling of the mouse wheel on the page to a certain coordinate. (there is no possibility to specify the scrolling speed).
​

Keyboard - similar to working with a mouse, you can enter any text or press any keys, which the script will execute automatically. You only need to specify the initial parameters.

For example, to enter an account login and password, you just need to provide the robot with the parameters and it will perform the entry by itself.

Data is one of the most important blocks.

Most of the actions in the browser involve interacting with data on a web page. For example, you may want to save data from all cards and then automatically substitute it into each account. Or download creatives for an advertising campaign from a specific folder on your computer. Perhaps you need to download a funny meme and send it to your grandmother 😁. By specifying the folder from where creatives will be uploaded to each ad cabinet, you can automate this process using our functionality. Detailed instructions on the correct use of actions from this block will be given in the future.

This block is designed to interact with the Dolphin {anty} functions on the outside of the profile.

Other - this includes actions that do not fall into the previous categories. You can add data processing loops or randomize bot actions (e.g., insert pauses) to make the script look as natural as possible, mimicking the behavior of a real user.

What Foreach, While and For loops are and how they differ can be found here.

After analyzing all actions, it is worth showing how to build your first small script and run it on several profiles at once.

As an example, let's take a look at the clicker for bitcoin faucet https://freebitco.in/. So far, all that is required from the script is to click on the Roll button once an hour and your account will be credited with satoshis. In this case you don't need to pass captcha, the button is just available for activation once an hour.

https://freebitco.in/

2. Connect the first block to the green Start circle.

You can run the script either on an "empty" browser copy or select one of your profiles from the list.

4. When you get to the site, being unauthorized, after a couple of seconds a pop-up window appears - which you can get rid of if you left-click anywhere outside the pop-up window. Let's add a Pause action and then a Click action.

To make a click, you can select the "left button" and "coordinates" option. In this case you can enter random coordinates (Y150 X150), the main thing is that the click was made outside the popup window and it disappeared.

You can determine the coordinates yourself using third-party extensions.

5. The pop-up window is gone, and now you need to authorize. Or you can be already authorized on this site, in which case you can go straight to the next item.

You need to click on the email field, enter your login/email there, then click on the password field, enter your password, and then click on the authorization button. In total, there are 5 blocks of actions.

In order for the robot to click a certain element, you need to specify which one. It may happen that the element will not always be in the same place, so coordinates will not help. We will use the "xpath" selector - you can copy it from the browser console in advance.
​

Copy the Xpath of the element into the "click" action in the constructor, and it will click exactly on that field.

The field is activated and we need to enter our mail. Select the Print text action and enter the mail in it.

Do the same with the password field. We select the Xpath element of the Sign Up button, and get to the main page.

Also, the Xpath element may not always be permanent - alternatively, you can use the CSS or Text option (specifies the element that contains the entered text). All this can also be seen through the browser console.

6. Now to the main page. All you need to do is click the Roll button, and the desired action will be performed. Select the"Click" block again, copy the Xpath of the Roll button, combine it with the previous action and done - the button is pressed.

You can automate many more actions in this way - there are probably other similar cases. But if you have any questions, you can always contact our technical support on the website\in the app or via Telegram @dolphinsupport_bot.