Trivia Generator

Curated by Skill Harbor: a trivia question generator by @mattdot (Matt Dotson), from his mattdot/trivia repository (MIT), listed here with credit to its creator. Give your Muse a topic, a single word, or a list, and the skill builds a set of hard questions whose difficulty comes from reasoning rather than obscure facts. Every answer is drawn from a set a player can actually walk through in their head, such as countries, U.S. states, chemical elements, league teams, or event host cities, sized roughly 30 to 100 items, and the conditions in each question prune that set down to a few plausible finalists a sharp player can bet on. Questions are written in plain, roughly 6th-grade wording, each answer is checked against cited sources, and every set opens with a short plan showing its dimensions, with at least 75 percent of the questions sitting on a dimension no other question in the set uses. A bundled catalog (answer-domains.md) lists the right-sized answer sets and the patterns that route almost any subject onto one of them. A set runs to 30 questions unless you ask for another number, and each question comes with its answer, a note on why it is interesting carrying a real figure, a follow-on comparison, and its source. A kid mode simplifies the subjects and the wording, topics can be mixed in one set, and any set can be rendered as a single self-contained HTML page for your phone from the bundled field-guide template: each question is a large tap target that reveals only its own answer, and the page plays without JavaScript. Honest caveats: this is a young, small project (created in June 2026, no stars yet), and it was first built for Claude Code. The original repository also carries a cross-game library, local Python scripts that keep questions from ever repeating between games, but that library lives outside the skill folder and needs the full local repo, so it is not part of what this listing installs and Muse will not keep that no-repeat promise on its own. What carries over in Muse is the core: generating sets, playing them in chat, and rendering the HTML version. The quality bar is demanding by design, so your agent may decline a topic that cannot support fair, rankable questions and suggest a closer one instead. Skill Harbor never reviews the code, review it yourself before use.