There's a lot of variables that could contribute to that -- how long is your game? how many other characters are playable? how skill-intensive do you imagine your battle system to be? what is the average level you expect the player to reach by the end of the game? etc. Because of all these elements, it's very hard to give a set number as a blanket statement.
I guess that strongly depends on how your battle system works and how long the game is. Getting 20 new abilities in the first 10 Minutes would be a bit overwhelming. But new abilities always make the game more interesting if they make sense for new types of enemies
Probably it depends on the game. I initially consider 16-20 skills for each character, with some characters having less than 10 skills. But recently thought about some kind of "less is more" thing and maybe no scrolling could be better. I think it might be more interesting to design the skills well, with their mechanics that makes the player think, like with @Milennin's game perhaps.
I've played games with large skill collections, and all the characters ended up behaving the same way: warriors tended to use 10 skills out of the lot, wizards went on 15 to 20 skills, accounting for self-buffs, summoning skills, changes of stances, and other things you use once or twice per battle.
all characters had more than 30 skills to choose from, but their final selection at any given time was less than 20.
this could be due to how many games have a quick bar to put skills on, and that is tied to the keyboard, and keyboards have only 10 to 12 number keys to manage that.
also, in many games, MP management is a thing, so having multiple levels of the same end effect using different amounts of MP each, sometimes is better than having a short selection of skills that get upgraded as the gameplay progresses.
if you're a warrior that relies on one heavy strike every now and then, you're fine... but if you're a wizard that has the power to hit several targets at once with one skill and you find yourself fighting one single target, you have to be able to use a single attack, and that often comes in the form of old skills that didn't get an upgrade and remained as they were in your 1st level.
this came in handy one time, where I was able to support a friend by spamming quick low-level healing skills that many other players would deprecate.
a 2v2 fight, warrior and healer on both sides, I had a macro of fast heal, while the other healer relied on physical shielding for his partner. THEY were stronger overall (by class), but WE outperformed them with smarts.
It depends on what skills do in your game. You don't want to flood the menu with skills the player will never use, but you also generally don't want them to select the same 3 skills in every fight. I think the most important factors are: 1. What kinds of challenges do you have to test the player in battle? 2. How does your character progression/build system work?
Skills should serve a purpose, so your combat system should present a challenge that the player must use skills to answer. In the Persona series there is a weakness system where the player gains an extra turn for hitting an enemy's weakness, so the player needs to have access to all of the different elements in order to find them. In the Tales series there is a combo system where the player creates their own combo attack by stringing different skills together so the player needs to have many attacks with different properties to find combinations they find satisfying to use together.
Skills should also play a part in character progression/building. The way the player gains stats, skills, passive abilities, etc. should all work together. If your game has a class/job system, then skills should be gained from that and there needs to be enough skills to make each class/job feel valuable.
Definitely get rid of any skills that seem useless 99% of the time. It's never good to clutter the menu with skills that don't have a real purpose in the game. Once you cut it down to the absolute bare minimum needed to make your combat work, then consider what skills might make battles better and how the player should get them.
As others said, it really boils down to your game. The number of actors and playtime, number of actors at a time, the skill systems themselves (Learning skills from level up? buying them? Can they upgrade into better skills? Are all skills kept forever or can they be switched out/changed over time?). And is it skills in total, if there is a way to switch, or skills on the actor at one time?
At one time, 10-12 distinct skills on each character at a time is the upper limit for games with 12 or less total party members. If your game has above that, I think dividing skills amongst them and shrinking the number of skills on a member at a time becomes more necessary.
Total is a different story. Really depends on your system and if that many skill options are necessary.
I am kind of same situation with skills myself here Plane . Doing my first vx ace project trying to make character class from scratch and skills for it. Planned priest, paladin and mage classes in party so total of 4 members. All are learned from lvl up so very basic since i am newbie beginner . I have mmm lightning, wind and water attack + sword techniques and dark techniques planned so maybe 20 ? not sure if i actually can balance them work nicely with other classes XD.
Basileus and Escalus especially made me think each of them how they fit my current prologue and yeah...dark cure and revive + water attack spells kinda were useless or not nesessary. So im down with 4 sword techniques 5 dark techniques and 5 spells so 14 i think. I am still unsure the balance but now i have starting point with battles.
I just successfully tested having events check if specific actors are in a particular party slot and then changing image to match. This will let me make scenes take into account who is in the party at the time and arrange them accordingly - this is important since I'll have extra actors beyond current party. The script used is the same from MV and is compatible with my current core plugin setup in MZ. Thank goodness.