Welcome, @rovaldes !
Here are some notes for you to consider.
If the original remote didn’t have a discrete ON and separate OFF button, then you’ve got to deal with the POWER TOGGLE ONLY command syndrome.
Bond has an option, called Trust State, which has a pretty big win / success rate if and only if you commit to using Bond, or BOND API integrations, for all control (as in, hide the original remotes and never use them).
I discussed this a little more in depth here:
… the ability to track state inside of the Bond system. What that means in reality is an 85-99% success rate of asking via API or voice assistant, etc a discrete “Light On” or “Light Off” command (for which Bond just sends a ‘toggle light’ command after verifying that the Bond-tracked last known state was the opposite) and getting the expected result – IF AND ONLY IF the original RF OEM remote is no longer used.
You have to commit to only using Bond app and any API-driven integrations ONLY to get any benefit out of the “trust state” feature.
How it works for me?
I use an ISY home automation controller to handle all my Bond devices whenever I am not using the Bond app. I enable trust state tracking in Bond for each device with a toggle.
Through that ISY, I can then send Light On and Light Off commands with a single tap on the Light On button or Light Off button of a in-wall or remote keypad (Insteon based in my case, although I even have a Harmony hack to allow home control buttons on those remotes to function) - a double tap on the respective Light On/Off will patch the state belief of the Bond’s tracking and then send the corresponding command to Bond again.
The double tap to PATCH Bond’s state belief is what saves me in the ~1-15% failure rate of something being delayed and therefore not actually turned On or Off as expected, or some guest finding the OEM remotes which I hide in drawers, or something else.
And there’s also this thread:
Track state changes when using remote