I wrote a crude note reader app that you can trigger with voice command. I'm assuming something like this is generally allowed in runs... right?

link: https://speedaudio.netlify.app

Example note is like this:

> Segment 8
	- change gc limit to 7 stacks
	- give labs to robots
	- restart robots, limit to 5 stacks
        # (TODO: not enough robots, start them earlier?)
		# TODO: fix gc belt to rc need more than 1 YB due to modules
	- stop rc to blue science and remove rc buffer
		# TODO: redesign so rc buffer can feed blue science, then rc buffer can stay
	- take modules and upgrade them to red chests
	- apply prod modules on rc, labs and gears.
		* Don't forget the additional iron belt for red green science
	- take some rails and dump grey assemblers
	- change wood chest to red chest in power pole factory

> Home Station
	- craft 4 stop and 100 signals (5 stacks of iron plates)
	- drag fuel belt to home station
	- build home station
		* Don't forget to blueprint the balancer and split the fuel belt
	- count modules and start speed modules
		* 870 speed modules, shoot for 900 prod modules
		# 610 prod modules, but already have 271 = 881?
	- delete forest & cliffs
	- stop flameturrets
	- upgrade gear and pipes to blue assemblers and apply prod modules
		* don't forget extra inserters for pipes
	- buffer blue science and restart it
	- apply prod module blue science and labs
	- correct blue science belt
	- apply prod module on frf and one plastic
	- collect robots for outpost, reconnect and count robots
		* T smaller than 700
	- buffer 28 stacks of FRF plus on the belt
	- get 48 prod modules

The note will be read line by line. It is up to you how detailed/brief the notes are. Indentation is ignored. Use the following four symbols to format the notes:

> Segment: each segment contain multiple steps (empty segment is not allowed)
- Step: each step may contain one "Additional note" (sound cue)
* Additional notes: additional notes will only be read on request
# Comment line: comment lines must be on its own line. They will never be read.

If you try to use this, please note:

• make sure to choose the correct mic in the browser settings (e.g. in chrome there is a mic icon in the address bar)
• to issue a voice command, be quiet for 1-2 seconds, then say the command, then be quiet for 1-2 seconds again (it is basically the web api speech recognition supported by the browser. But this is also convenient, I have been talking freely without ever accidentally issuing any command)
• before using, try saying a couple things and see the percentages shown at the bottom to check whether it recognizes your words accurately (if any of the guesses matches a voice command, e.g. "next one slow", it will be executed)
• there is one useful command I forgot to list there: "repeat it"
• i've used this for 2 years but it has never been tested by anyone else. It defaults to "Google US English" - I assume every browser will have it but didn't test

code (angular): https://github.com/kevincentius/speedaudio

This is interesting. I will pass it along on discord. As for now i can say that there is not a rule about it yet and currently it seems okay to use. Are you perhaps on discord and / or part of the steelaxe discord? If so, what is your username, so i can add you to the contact thread if you require additional clarification of what is or isnt allowed: that being not interacting with the game or reading the memory.

Thanks! Yeah, I'm Kevin (kevincentius) in there.