Built for people who bet on races, not tickers
Taonets started with a simple irritation: the chain tells you WHAT happened, GitHub tells you HOW it's built, and neither tells you whether you should care.
What we track that others don't
- • Current mission. Every subnet's on-chain registered repo — description, topics and README — classified into a mission type from the project's own words.
- • Emission dominance. Who actually captures each subnet's payout: top-uid and top-10 concentration, earning-uid counts and active race size, read from the chain's Emission and Active maps.
- • Integrity. A transparent chain-evidence score: emission concentration, repo dormancy and emission status — every subscore cites the storage read behind it.
- • Yield. Live registration burn vs top-slot payout vs pool depth — entry economics before you spend a TAO.
- • Honesty. Where a signal would need data we don't ingest yet (metagraph history, transfer graphs), we say so instead of simulating one.
How to read this deployment
This site reads live Bittensor chain data: our worker reads subtensor storage directly (owners, identities, emissions, pool reserves, burns, active miners) and each subnet's registered GitHub repo. Numbers are the chain's own; semantic labels come from each project's self-description. If the worker hasn't synced here yet you'll see a banner and placeholders — never fabricated stats.
Stack
for the curious
Next.js 15 (RSC-first), Tailwind v4, motion, React Flow, ECharts, Drizzle/Postgres and a worker fleet reading subtensor over WebSocket — packaged as a pnpm monorepo with the scoring engines as a pure, fully-tested TypeScript library.