This game was entered in Spring Thing 25, and it was a pleasure to play. It consists of three smaller games all tied together by a larger meta-story.
The first game I played was Sticks and Stones, which reminded me a lot of Tavern Crawler, an IFComp winner from a few years ago. This mini-game is played completely by arrow keys, with arrows serving both as navigation and as ways to interact. It includes combat, a money system, two protagonists, a variety of NPCs, and made me feel like I was playing a really enjoyable game (as Spring Thing organizer, I can reveal that this game actually came in 3rd in votes, and was only off by 1 vote).
I then played Treasures of the Deep, which I had seen someone else mention as short and linear, which it was. I saw the author mention in his postmortem that he went out of this comfort zone for the writing here, and I think it paid off. It gave the feeling of 'this is a person who often writes gripping stories' rather than 'this is this author's first time.'
I then played The Labyrinth, which I enjoyed but didn't like quite as much. The escape room model matches his other games, but the difficulty is reduced, so some of it felt like busy work (this is amplified by me playing all 4 people at once. I've only ever managed to pull off one multiplayer IF game.) I saw in his postmortem that new players got confused, so it wasn't super easy, but I think the real issue here is that it's important in a puzzly game to model early on the type of content and puzzles you're going to have. The Labyrinth has a lot of frustrating things (including early deaths) and content you can't interact with yet (like door riddles and the patterns on the well) before you reach the first sharable piece of information. I think it might have been better to have an optional early piece of information you can share that just gives the other players a thumbs up or encouragement or something so they can get the pattern down early. But I'm not sure.
I enjoyed the meta-puzzle at the end. I had expected it to be more subtle and had copied down patterns of gem-pushing in each game, sure that I'd have to do something like navigate to the menu to find a secret meta-puzzle by the order I push gems at the beginning. Instead, everything was quite clear and out in the open, which was fun.
Seeing only one review on IFDB, I assumed very few people had completed the game. I was really surprised to get to the end and see that over 30 people had recorded their names. I guess that really shows that the general rule of the internet applies (10 times as many people see something as interact, 10 times as many people interact as comment).