For board game designers

Find the questions your players will ask, before they ask them.

FAQproof reads your rulebook and card list the way your players will, beginners and veterans alike, and highlights the rules questions likely to come up at the table. Each one points to the page or card it comes from. We don’t rewrite your rules; we show you where they’ll trip players up.

Join the free beta How it works

Tested on more than a dozen published games, against the questions real players asked after release.

Flags · draft 33 of 38
Timing Can you use a “once per round” ability on the same turn you gain it? p. 6 · Abilities
Term Card 42 says “exhaust”, but the rulebook never defines it. Card 42 · Glossary, p. 14
Conflict Setup gives each player 5 coins, but the solo rules say 4. Which is it? p. 3 vs. Solo rules p. 1
Illustrative example from a fictional game.

How it works

A careful reader for every draft

Blind playtests are precious. FAQproof catches the confusing points before you run one, so your testers spend their time on the game, not on questions the rulebook should have answered.

It isn’t a single prompt to a chatbot. We built a system for rulebooks: it reads your rules several times over, goes through every card one by one against the rules, looks for answers that are in the book but easy to miss, and groups everything by page so you can fix each question where it lives.

  1. Upload your draft

    Your rulebook PDF and your card spreadsheet, plus any solo rules or appendix. Rough drafts are welcome.

  2. Get the questions

    The questions players are likely to ask, grouped by page and by card, each linked to the exact place it comes from.

  3. Triage and fix

    Mark each flag “will fix”, “not a problem” or “intended”, then take your to-do list back to your rulebook.

What we highlight

Every question is tagged with its type, so you can see at a glance what kind of fix it needs.

  • Timing and order“Does this happen before or after scoring?”
  • Edge casesTies, empty decks, running out of pieces.
  • Undefined termsWords on cards that the rulebook never explains.
  • ContradictionsBetween pages, or between the main rules and the solo rules or appendix.
  • Card–rule interactionsA card effect the rules don’t cover, or a rule that changes what a card means.
  • Card data errorsOverlapping score brackets, duplicate card names, icons that are never explained.

How it looks

Your rulebook, with the questions marked

Screens from a report on Tidepool Couriers, a small game we invented for this demo.

The rulebook view: page 2 of the demo rulebook with underlined passages and numbered questions in the margin; the selected question asks whether a player may take the same action twice when only one is possible.
Each question is marked on the page it comes from. Pick one to see why players will ask it and the words that cause it.
The cards view: a grid of demo cards with their open questions; the selected card's question says the card and the appendix give different carrying limits.
Every card is checked against the rules, so you see which cards will raise questions and why.
Join the free beta
We flag. We never write.

We raise questions and show you where they come from. We don’t rewrite your rules or suggest new text, so your rulebook stays in your own words and your own voice.

You decide what’s a real problem. Some “questions” are exactly the choices you meant players to make.

Privacy

Your game stays yours

You’re trusting us with a game that isn’t out yet. Here’s what we promise.

Private by default

Only you can see your files and your report. There are no public links.

Never used for training

The analysis runs on Claude through Anthropic’s commercial API, which doesn’t train on your files. Neither do we.

Delete anytime

One click removes your files, the text we extracted and your report from our systems.

Honest about AI

FAQproof can miss things or get them wrong. Treat every flag as a question to check, not a verdict.

Read the full privacy page →

Who’s behind it
PhD in machine learning · 20+ years in AI · lifelong rulebook reader

I’ve spent more than 20 years in AI. At home, it’s board games.

I did my PhD in machine learning, back when that’s what it was called, before everyone started saying “AI”. I’ve worked in the field ever since.

My partner and I have played board games with our kids since they were little, and we still do now that some of them are grown up. Our shelves are packed. I’m the one who reads the rulebook, and I love it: learning a new game is half the fun. For some reason I remember every nitty-gritty rule, even for games we haven’t opened in years.

FAQproof is where those two meet. It reads your rulebook the way a careful new player would, so the questions land on your desk instead of at someone else’s table.

Join the free beta

I’m looking for a small group of designers who are one to four months from a Kickstarter or a publisher pitch.

  • A free report on your game
  • Your files stay private
  • A 30-minute call to hear what you think
Email hello@faqproof.com

Tell me a little about your game. Please don’t attach your files; I’ll send you an invite to upload them yourself.