This repository's root is a finished static export (plain HTML, CSS and JS, no build step, no backend, relative paths, it works from any directory): index.html, how-it-works.html, fees.html, token.html, modules.html, contracts.html, ai.html, faq.html, 404.html, llms.txt, llms-full.txt, protocol.json, hashmap.json, vp-icons.css, README.md, brand/ (4 icons) and assets/ (25 files: page scripts, chunks, one stylesheet, fonts) at the repo root, 44 files in total. The site root is the commit root: every file is hosted exactly as committed, none renamed, moved or edited. Check its content, then pin and name it; do not build, change, rewrite, minify or add anything.

Fixed parameters, fail, don't change: repository root as is, commit 2aec7914338c2ff5631a6c21341c87cfd0650783, site label docsmoneycanbuyfun, 44 files. If any of them cannot be used exactly as given, stop and report; do not pick another label, commit or file set.

What it is: the documentation of moneycanbuy.fun, paid dares on Ethereum mainnet judged by identity.md NFT holders and the IMD swarm. Eight pages: overview, how it works, fees and money flows, the $YOLO token, the modules (trophies, predict, sweetener, hidden list), the mainnet contract addresses and functions, a guide for AI agents, and an FAQ. llms.txt, llms-full.txt and protocol.json are plain-text and JSON copies of the same facts for machines.

Accepted by design, not findings: the site loads nothing from the network (fonts, icons and the search index are local files); search runs in the browser only; the outbound links (the app's site under sites.imd.fun, the project's X and Telegram accounts, Etherscan addresses) are plain anchors a visitor may click; the code samples on the AI agents page are text for readers and are never run by the site; there are no forms, cookies, accounts, analytics or wallet connections.