@Joan-Canario, you should be able to use selenium to do that. What's the website? I can probably do an example. I might use this: https://automatetheboringstuff.com/chapter11/

Awsome, @Călin-Cătălin-Alexandru I'm glad you figured it out.

@Maximiliano-Farinaro, can you confirm your Windows version is 64bit Windows 10? I've seen problems like this on 32bit Windows.

@VRMan3D, that's funny about using "shutdown". Yea unfortunately Alexa reserves certain words like "stop" and apparently "shutdown". What word are you using for your "Voice" word for your command? Alexa understands certain words better than others. It often helps to look at your voice command history at https://alexa.amazon.com/ to see what she heard. You can compare your voice word to what Alexa hears to pick something she hears well every time. Another option is to use the TRIGGERcmd Smart Home skill. You say, "Alexa, turn on notepad" instead of "Alexa, ask TRIGGERcmd to run notepad." The Smart Home skill works better because Alexa knows your commands ahead of time and only has to match what you said to one of your commands.

@Paulo-Eduardo, awesome. Nice job. I'm glad you got it working.

@Raphael-Neves, yes, if you subscribe you can run more than 1 command per minute.

@eggbean, man that is brilliant! Nice job. I can't wait to read that blog post. I might want to do something similar. You don't need npm to run the agent. It's useful for installing the dependencies but it makes sense to remove it after. Thanks for sharing.

@Russ [image: 1594257418444-6aaf83bd-9770-4de5-a697-781d7b5ba91f-image.png]

@Alcir-Junior, awesome.

@RAFFY, I'm finding I get the same error when I try to create an IFTTT applet. I confirmed I haven't changed any code on my end. I'm looking into it. Thank you for reporting this problem! EDIT: I got a response from IFTTT that wasn't very helpful. I sent them another email and I'm still waiting. Basically, right now you can't create new IFTTT applets that use ingredients in the params field. For now, you could use the old Webhooks method. EDIT2: It's fixed now.

@Furkan-Cevik , I got your chat. Thanks for confirming it's working now, and thanks for reporting the problem.

@eggbean, I turned off emails from the forum because I was getting a lot of spammers. Since then they've stopped posting, so I'll turn it back on. To add/remove commands, edit your /root/.TRIGGERcmdData/commands.json file. Make sure you make them background commands because you're running the background agent. The background agent adds, removes, and runs background commands. FYI, if you find that your commands.json keeps reverting as soon as you save it, that means you've made some mistake in the json formatting so the agent restored the last good version of it.

@Josiah-Richards and @jgodney, please be aware, there's an official Integromat integration now.

@frederico-allan, I wasn't able to program the foreground agent on Linux to startup automatically after login. I couldn't find a standardized way to do that. If you know of a way, please let me know. I'm glad you got it working.

@Marcelo-Botelho, you don't need to use the computer_on_off.sh. That's a complicated scenario, so you should not start there. You have two computers, and you're trying to use one computer to turn on the other computer, correct? My video shows me using a Raspberry Pi to turn on my Windows computer, but you don't have to use a Raspberry Pi. If you have two Windows computers, you could use this utility instead: https://www.nirsoft.net/utils/wake_on_lan.html WakeMeOnLan.exe has a spanish version, and a GUI which can make it easy to test, then after you get that working, you can make a TRIGGERcmd command like this: c:\tools\WakeMeOnLan.exe /wakeup 40-65-81-A7-16-23 This assumes you copied the WakeMeOnLan.exe utility into t:\tools and 40-65-81-A7-16-23 is the mac address of the nic in the computer you're turning on. You can find the mac address by typing ipconfig /all on the computer you want to turn on. You'll get output like this: Ethernet adapter Ethernet: Connection-specific DNS Suffix . : lan Description . . . . . . . . . . . : Realtek PCIe GbE Family Controller Physical Address. . . . . . . . . : **C4-65-26-2A-93-97** ...

@Ivan-Freitas, I posted a mac solution here in the mac section.

@Joe, 1 - Only Foreground Triggers work in a user's regular account - because the background service only runs as root and uses /root/.TRIGGERcmdData. 2 - Root can support both Foreground and Background Triggers - because you can install the background service (see #1), and because you can login as root and run the agent in foreground mode. 3 - But running Triggers as Root will mask/override all the Triggers in your regular account - I wouldn't say mask/override, it's just separate as you covered in #4. Running the agent as root will create another separate computer entry in your account associated with /root/.TRIGGERcmdData. That computer will be named the same initially, but you could rename it. 4 - The exception to (3) is if you can instantiate two "Computers" -- one pointing to Root for the "Background", and a separate "Computer" pointing to your normal Linux account to support the "Foreground" - correct, although as you pointed out in #2, you could also run the foreground agent as root. You create a separate computer just by running the agent as the other user. If the ~/.TRIGGERcmdData directory doesn't exist, the agent will prompt you for a token, create a computer record in your account, and store the ID for that computer in the computerid.cfg file. Different Linux accounts could setup different computer records, or you could give everyone sudo access to be able to edit /root/.TRIGGERcmdData/commands.json to run background command via the background agent. The agent stores the .TRIGGERcmdData directory in the home folder of the user running the agent. The files in that directory are the key point of connection to the mother ship. Good questions @Joe.

@Michael-Murphy, thank you for reporting this. I was actually worried people were getting that email when they shouldn't, but you're only the second first person to report it, so thank you! If thought the first person's issue was a one-off, but apparently not. I think fixed it just now, but anyway I decided to make it an opt-in feature instead of emailing by default. Now you'll see this checkbox in your profile, and it's unchecked by default: [image: 1590713174710-0583d88f-2e1c-4942-9e9d-3e26a020fe30-image.png]

Awesome @mikeraineri.

@mikey186, thanks. I'll fix that. EDIT: Fixed.