> as I mentioned earlier, I was working on a tool to bridge speech recognition
> from Windows VM to drive a Linux environment. I now have something good
> enough for plaintext dictation. You can find it on git-hub
> https://github.com/alsoeric/speechbridge
Now that I've looked at your project a little closer, I see you are using python. At-spi has python bindings that you could use to query window/focus state etc, and Orca being in python is likely your best reference.