Real-time Oracle Intelligence Platform + Paper Prediction Games

Pyth Community Hackathon
1

DeltaScope

Team: @0xPilotSB
Submitted: 29 March, 2026

Answer Capsule

DeltaScope is a real-time oracle intelligence platform + paper prediction game that monitors the gap between Pyth Network oracle prices and Hyperliquid DEX mark prices. It uses triple-source Pyth ingestion (dual Hermes WebSocket + REST polling) with Pyth Lazer ready for sub-50ms updates, an AI assistant with 6 Pyth-powered tools for querying 1,930+ price feeds, and a Predict & Win game where users bet on price direction using live Pyth oracle data — with confidence-aware settlement that refunds when oracle data is unreliable.

What It Does

Oracle prices and DEX mark prices should match — but they don’t. The gap reveals liquidation risk, funding rate mechanics, and infrastructure health that most traders are blind to.

DeltaScope gives traders real-time visibility into:

  • Oracle-mark spread — Real-time discrepancy between Pyth oracle prices and Hyperliquid mark prices across 8 major assets
  • Pyth publish delay analytics — How stale is your oracle data right now? Median publish delay tracked in a 10-min rolling buffer
  • Infrastructure latency monitoring — Pyth Oracle Delay, Hyperliquid REST API latency, WebSocket delivery, and overall health score
  • Predict & Win — Paper prediction game where users predict price direction (UP/DOWN) using live Pyth prices, with alarm-based settlement, streak bonuses, and a global leaderboard
  • AI-powered analysis — Natural language queries across 1,930+ Pyth price feeds with 6 structured tools

This is the layer beneath the prices that nobody else shows.

Pyth Features Used

Check all that apply:

  • Price Feeds (on-chain or off-chain)
  • Entropy (randomness)
  • Both

Links

Screenshots / Media

Screenshot 2026-03-28 at 20.40.20
Screenshot 2026-03-28 at 20.40.201920×1150 194 KB

Screenshot 2026-03-28 at 20.40.48
Screenshot 2026-03-28 at 20.40.481920×1039 239 KB

Screenshot 2026-03-28 at 20.42.01
Screenshot 2026-03-28 at 20.42.011920×882 209 KB

![Screenshot 2026-03-28 at 20.42.39|690x375]
(upload://dRwFq2Z4Fl0fjrRH5sAk34ormrz.jpeg)
Screenshot 2026-03-28 at 20.43.32
Screenshot 2026-03-28 at 20.43.322936×1592 459 KB

Screenshot 2026-03-28 at 20.44.59
Screenshot 2026-03-28 at 20.44.591920×1051 180 KB

![Screenshot 2026-03-28 at 20.45.33|690x372]
(upload://mh1ouNoiA4qiKf7PLIV0ZkTgyJ8.png)
Screenshot 2026-03-28 at 20.46.06
Screenshot 2026-03-28 at 20.46.062936×1604 306 KB

![Screenshot 2026-03-28 at 20.46.42|690x372]
(upload://eBQki0AfCc0A2huzz5424h3xy2o.png)

Tech Stack

  • Framework/Language: React Router 7 (TypeScript), Vite, Tailwind CSS 4, shadcn/ui

  • Blockchain (if applicable): Hyperliquid (Not Yet Deployed, Still ongoing on my mind usecase)

  • Agent Framework (if applicable): Vercel AI SDK + Cloudflare Workers AI (6 structured Pyth tools with codemode)

  • Deployment: Cloudflare Workers + 4 Durable Objects (edge-deployed, 24/7 uptime via DO alarms, Smart Placement

Content Contributions

Licensing

This project is licensed under Apache 2.0

Eligibility Confirmation

  • I am 18+ years old
  • I am not located in an OFAC-sanctioned jurisdiction
  • I confirm this is an original work created during the hackathon period
  • I have read and agree to the Terms & Conditions