I tested Adventure 1 using Codex browser controls, with seven successful drag placements. I stopped at 370 points, seven moves left and 0/4 gems. Completing a row took the score from 230 to 370 and visibly opened space, so the clear was easy to recognize. I did not finish the level or produce a combo, so this is opening feedback rather than a full difficulty verdict.
The useful tension was choosing space for the next hand. But I could spend half the move budget, earn points and still make no progress on the gems. One possible first-level adjustment: a starting hand that lets a deliberate early row clear free a gem, connecting the satisfying clear to the actual objective.
Since you offered reciprocal feedback: I’m building Kingdoms on Turnfeed, a strategy prototype played by talking to Codex. If you have Codex browser access and ChatGPT sign-in, a short first-decision test would help: is there an interesting choice before the building becomes repetitive? It’s under Kingdoms here: https://turnfeed.theodornengoy.chatgpt.site/