Takar

Each machine below had a budget for intelligence. Before moving, it bought the checks its task needed through Telegraph's ranked market, compared them with the nearest airport weather station, and decided whether to go.

Simulated fleet, real purchases. The units are illustrative. Every purchase settled on Ethereum Sepolia testnet with test USDC, and every station reading is a published METAR report.

Loading the receipts…

What the receipts showed

  1. The top-ranked storm miner ignores the place you ask about. Storm checks for Makassar, Palangka Raya and Bandung were routed to miner 37007, which answered every one with Tropical Cyclone AKVAN-26 in India. Once, that answer was shaped into a 296 km/h gust for Bandung. Reported with receipts in bounties#211.
  2. A storm answer can be paid for and empty. The next-ranked miner answered Palangka Raya with every field blank, so the drone there held.
  3. A station can disagree with a recorded-weather answer. A miner gave Makassar a 31.9 °C high for 15 November; the Hasanuddin airport station recorded 34 °C across 49 reports.
  4. When two independent sources agree, the machine goes. The Makassar drone flew after the storm miner forecast a 22.3 km/h gust and the airport station measured 17 km/h.

Forecast scorecard (Mission 13)

Loading the scorecard…

Show every forecast

Conditional payments (Missions 14 and 06)

Test USDC locked in TakarSettle against a condition, settled once Telegraph answers. If the condition is met the payee is paid, otherwise the payer is refunded. Each settlement carries the signal it rests on.

Loading…

Every purchase

Show the purchase ledger