• Draw the game - in a graphics programme or on paper, but draw it to scale.  That way you can see the layout you have in mind and see if everything will fit.
• You need only the main ideas to start with, and the drawing. 
  • What's going to be unique about the game?
  • What's going to be the fun aspects?
  • How do you win?
  • How do you lose?
  • Will there be achievements or rewards?
  • What happens when you die / lose?

• Decide on your target devices (i.e. pc or mobile etc.) and keep that in mind when you plan out the size and extra effects.
• Create a small prototype of the game just using basic shapes - sometimes a game sounds fun in theory, but isn't when it's actually written. It will also give you ideas of what works, what isn't going to work, what is going to be a problem to program or to create graphics for etc. When I have the game working on a template level and I can see that it's viable - that's when I start planning the detail and the things to make it look pretty.  I've also learnt that when a game is extremely boring to develop that it is also going to be extremely boring to play - I abandon those or rethink the ideas.
• Make a list of all the global variables and data structures that you will need (for example, score, health, weapons.  But don't go into too much detail - just the basics - it's enough to know you need to plan out the weapons in a data structure without necessarily listing every weapon you can think of.  You can add those in as you're working on the game and most software is flexible enough that a list can grow or shrink without much effort. 
• Note down all the interface / menu items you need and decide where and how that is going to be implemented - will the buttons always be on screen or will you have a pop-up menu?  It's important to create with the pause / music / sound options in mind from the start, because they should be settings where the user can choose and you need to plan for that when you code - i.e. you can't just say, play sound, you have to say - if user wants sound - play sound.
• I probably have at least 20 games written down and planned in detail that I've never written.  When I'm really excited about a game idea, I tend to plan them a lot less and jump into creating them sooner.  But that might just be me...  

Trello is great! If you're looking for a one-stop-shop for to-do lists, project management, document archiving (maybe for GDDs or story docs), web clippings and more - all with the option to share collaboratively - I can't recommend Notion enough. 

I switched from Trello to Notion for project management and production bits, and it's been a game changer. 

https://www.notion.so/

chez_dev, I'd suggest mixing it to start coding and add thoughts within its text or in another file. Personally I'd add info like: banner's design, game's title, in-game features, stuff that being edited but not working or just got fixed recently. Or publishing-date, additional scenes or backstories, etc.

It sounds rather boring just texting how to get its story appear without directly coding it, I guess. Before releasing I'm saving a copy without script's // and similar !--comments-- so it just includes the attributes without thoughts there.

I also fuse some parts, divide it, editing or adding paths, takes a while also if I loose interest in one section, getting it mentioned as hint & continuing on another region, e.g. less detailed or just into alternate endings, etc.

In my case it's more like no scripted version, just getting surprised how the game turns out before publishing it and coding while thinking of how it could appear for viewers.