You might also be able to do Harold Frumpkin hehe

Love seeing a show get a bunch of characters made into LoRAs. Various recommendations from a veteran creator that wants as many awesome LoRAs to be made and them to be the best they can be (note you should have gotten a decent chunk of blue buzz from me posting images using your LoRAs)
1: Don't use "givenname surname" triggers. These tend to have random parts of them read as independent tags (e.g., I had a LoRA where the "road_wargear" in the trigger was read as "road" and kept seeing roads spawn in that LoRA till I retrained it) and Civitai's generator has a filter that likes to reject these because it thinks you're trying to prompt for an actual person. It's a relatively small risk, but impossible to correct after training when it does become a problem and a totally avoidable issue. You can set a trigger as any arbitrary set of text, but my favorite is four single letters separated by underscores that combine the character name and series name (e.g., for this "A_G_D_D")
2: You should list the trigger and clothing tags used in the description (preferably in order of head to toe so stuff too low for the intended camera position can be easily removed when copy/pasted, e.g., not including "shoes" for a cowboy shot) rather than expect people get them from the sample pics. Having tags planned out and written down like this before hand also makes it way easier when tagging pre-training.
3: From my experience, for Illustrious you generally don't need to describe a character's body unless it's non-standard (e.g., some pics a character has much larger breasts than normal, or there's an episode where she gets super muscular), or you're trying to correct an imperfect LoRA at prompt time. Likewise it's enough to give each article of clothing one tag, preferably a more specific one Illustrious knows when possible (e.g., "t-shirt" instead of "shirt") because that makes it play nicer with other LoRAs (so they aren't trying to fight over the same stuff), though wearing methods (e.g., open jacket, jacket on shoulders, unworn jacket) should still be used (you technically don't need to tag clothes at all, but that means they can't be changed easily). Hair is about the same, though you do want to say hair color as well (because different works have very different pallets).
4: Just leave the strength settings blank (unless you know it needs a non-1 default strength). The site will default those to a reasonable -1 to positive 2 range with a default strength of 1. Using something like 0.80 strength is reasonably common when trying to get multiple LoRAs to play nice so no reason to make it impossible.
5: Assuming you're saving your tagged training data (You should!) I'd love to see that data all combined into a style when you're done. Once you've got a bunch of character stuff tagged, it's just a matter of adding a supplemental data pack (about as much data as a character or two) for broadening what that data includes (e.g., background/object focused shots, crowds, male characters) and training
6: Pony scores don't do anything in most Illustrious checkpoints, they're something Pony was built for. A few checkpoints do respond to them due to some shenanigans in how they were made, but for most it's just a waste of tokens (GPU attention span)
7: A LoRA's only shows up for a user if its main gallery has an image that matches what ratings they have visible. In practice the main thing to worry about is having a PG image of some form so it shows up for users with visibility of mature pics disabled AND those that aren't logged in. A prompt with "portrait" framing (and something like t-shirt if needed) is generally safe.