For an RPG Maker game that you're making by yourself, you don't need much of what a normal GDD will contain. Still, I find it much easier to change things when they are done in text than having to change something within RPG Maker itself after spending 2 weeks on it...nothing like needing to restructure the order of skills in the database after creating 50 of them...
Anyways, I'm a bit of an Excel maniac and I like to make spreadsheets for everything I do. Documenting has become habit for me. I create rough stat tables, tables that describe my skills/spells, and sometimes a rundown of the EXP the player will get and their total EXP on a per battle basis. I use Notepad++ for more "text" based recordings...such as what the high level design of my game is, the story, etc.
I'll use some specific examples from my game instead of talking in a "high level". The following is my process. Do not mistake this as a "How you should document your game". :)
Core Game Concept
My Notepad document starts off describing my core goal: "To translate the game structure of a Mega Man game into a turn-based RPG format." This is my CENTRAL idea - the whole reason I'm making my game. If I see a script or read a cool game mechanic, I first consult my central idea. Does this script have a place within my game? Will I be straying from my main idea in order to find a way to incorporate this script? Does this cool mechanic fit within the parameters of my game?" It is very easy for someone (especially someone like myself) to constantly come up with a new "idea" that you get excited about, and you become conflicted. Do I scrap my current project in order to start working on this new idea I'm super excited about? Or do I try and shoehorn this idea into my current game? If I don't have a core game concept that I can consult, I'm very likely to choose one of these two options. If I still find myself motivated to carry out my original game concept, then I can just pocket that new idea and save it for later.
Gameplay
In Mega Man, you have free reign to choose the order in which you play your levels. So I want to include this in my game. My game will be separated into "dungeons" connected by a central HUB. The core concept of Mega Man is that you earn a new weapon (ie: power) whenever you defeat a boss, and each boss is weak to a specific weapon. For those that play the game over and over again eventually determine which weapon is good against which boss, and then determine the ideal order in which you play the levels. So I'm going to take the idea of earning a new weapon after beating a level, and turn it into earning a new party member after beating a level. And just like a Mega Man boss is weak to a certain weapon, I'll want to design my bosses to be weak (ie: easier to defeat) against a certain party member.
So I need to make sure that my entire battle system and skill design revolves around this. I need to differentiate my party members enough that Character X truly is "good" against its intended boss.
Here is where I plan how I'm going to approach battles - whether they will be fast or slow, whether I want the characters to be customizable or confined to a role, etc. I feel the best way to make a party member feel "good" against a certain boss is to use the same battle mechanics in Final Fantasy X - that is, giving each of my character's a specialty that helps them take down a specific family of enemies/monsters. This is forming the CORE of my battles. I'm defining families of monsters that will have distinct characteristics and each of my party members are going to have a unique tool that helps them deal with a certain monster family. So I have my "Slime" family that behaves pretty much the same as the Slime enemies in FFX - they are highly resistant to physical attacks and usually have an elemental weakness, making them ideal targets for your Mage character. I need to be careful that a monster family is not too difficult (or worse: impossible) to defeat should you lack the character that's designed to take them down. I don't want to FORCE the player to use the Mage to defeat Slime enemies, but it sure will help if he knows an area has a lot of Slime enemies, that he should probably bring the Mage along.
Bosses
I have an entire section dedicated to my bosses. I have six levels, therefore six bosses. Not all are designed as of now, but I pretty much write up my own individual BDD (Boss Design Document) for them and just write as if I'm talking to someone else. I list the different ideas I have for designing a certain boss, and I'll mention what I don't like about each idea or why elements of each idea are in conflict with what I need to achieve with the boss. Example? My "Slime Boss" aka. "Mage Boss". This boss's REQUIREMENT is that he is easier to defeat if you have the Mage character (I don't name my characters here, names can change so I refer to them as their role). He must not be impossible to defeat if you don't have the Mage. So what does the Mage do that no one else does? The Mage is the only character in my game that has access to all four elements within a single kit (other characters might only have access to one or two elements). So I need my Slime Boss to be designed to be easier to defeat if you have a wide range of elements. So every "gimmick" or "mechanic" I'm thinking up of for my Slime Boss is run against my design requirements. If I use this mechanic, it doesn't really promote varied element usage: it's rewarding the player for spamming the same element, which is in direct opposition of what I want from this boss.".
Story
Story is not really the focus of my game. I don't even have an idea for what the story is going to be. I don't think it matters too much. It's not the focus of my game - if I can come up with something cool that fits with my game's structure, that'd be great, but I really haven't spent any time thinking of story yet. I'm more intent on prototyping my gameplay and making sure my party members interact with the enemies the way I intend them to.
Still, WHEN I start planning it, I'll probably start brainstorming some ideas in this same Notepad document.
Spreadsheets
The rest of my recording is done in Excel, where I can go crazy making spreadsheets. I have a Stat Tables for each of my player characters. I only create stat spreads up to Lv.10 as it's unnecessary to plan any further until I get more concrete gameplay created. Each skill/spell can be upgraded three times, so I have tables depicting the stats of each spell and how it improves with each upgrade. This is far easier to edit than having to change a slew of skills in RPG Maker's database. I have spreadsheets for my enemies, describing what differentiates Slime Enemy C from every other monster from the Slime family. I make sure that I don't have two Slime enemies that function exactly the same way. Either they have varied elemental weaknesses/resistances, cast different skills/spells, have different stat spreads, etc. But they ALL share the common trait of being resistant to physical attacks. That is a requirement for this family.
I have spreadsheets started for my weapons and armour, planning the possible ways the player will be able to prepare for each level.
And finally, I have spreadsheets for the EXP and JP gains for each and every fight for each level. I need this to balance out the overall EXP or JP you can receive from each level. I don't want the player choosing Level B first for the extra EXP it gives compared to the other 5 levels. So these EXP tables help me ensure that no matter which level you play first, assuming you fought all enemies, you'd be at 700 EXP give or take 5-10 EXP.
So far, that is everything I've documented for my game. It has helped me a ton, and it helps that I just love creating spreadsheets, so it's not boring for me to spend so much time in this planning stage. Still, if ever I start losing that spark, or motivation - I try to take a break from planning and start actually implementing. Seeing CONCRETE progress on my game is the best way for me to retain motivation.