OIN — Verifiable / Cryptographically-Timestamped Prediction-Market Intelligence with On-Chain Receipts

OIN tracks informed prediction-market wallets in real time: Transfer-Entropy identifies which public Polymarket wallets historically move before the crowd; you get the raw, hash-sealed flow the moment a tracked wallet takes a position — before it's reported elsewhere — and build your own model on it.

Every observation is hash-sealed and Bitcoin-anchored BEFORE any market outcome is known — verify any of it yourself at verify.html.

For the sharp who builds their own models and needs raw, early, KYC-free data with proof of when it existed. We sell verifiable receipts and lead-time — not predictions. Not for copy-traders.

373 provenance-sealed signals currently in the corpus · live count at /api/v1/signals

⛓ Bitcoin-anchored provenance · OpenTimestamps

The entire signal corpus is committed to Bitcoin daily via OpenTimestamps. Anchoring is performed by an independent third-party calendar service we cannot back-date. Confirmations land within ~24h. Verify any commit yourself →

What makes OIN different

Honest disclosure — read before buying

OIN does not sell predictions. Our directional accuracy is published daily at scorecard.json (auto-recomputed, on-chain ground truth) — read it before buying. Aggregate directional accuracy is not statistically distinguishable from the ~56.5% base rate (currently 63.1% on 271 resolved-and-graded on-chain calls, concentrated in one source — selection bias, not replicable); no demonstrated directional edge. We sell provenance, not forecasts. No win-rate claim, no performance promise.

What OIN sells is provenance and lead-time priority: cryptographically-proven observation timestamps you can join to on-chain settlement and build your own model on. Analytics & research only — not financial advice.

⚒ Dark Shipping signal (example)

Named shadow-fleet vessel positions in the Black Sea, timestamped and sealed. Cross-reference AIS gaps with geopolitical events. Raw feed, no interpretation supplied.

The proof hook — Iran, Jun 2026

OIN's corpus for 2026-06-01 was committed to Bitcoin in block 952139 on 2026-06-03 — seven days before the 2026-06-10 war-resumption headline. Every Iran observation below was hash-sealed before the event. Read the full timeline with verifiable receipts →


Get access

Payment in USDC · anonymous · no KYC · no account signup required. Start free with no key (top 10 of /api/v1/signals), upgrade when you want the full stream.

Analyst
$29
per 30 days · USDC · 120s delay
  • Full keyed /api/v1/signals
  • SHA-256 provenance on every signal
  • 120-second signal delay
  • 30-day lump-sum · no auto-renew
  • X-API-Key auth
Start via Telegram →
Firehose
$299
90 days · USDC · real-time · 5 seats max
  • Real-time raw firehose
  • 0-second delay · full provenance API
  • Every commit + timestamped seal
  • 90-day lump-sum · no auto-renew
  • 5 seats maximum
Start via Telegram →
Founder
$499
one-time · USDC · real-time · 1 of 5 seats
  • Everything in Firehose
  • Future data products included
  • Grandfathered at founding price for life
  • Direct ops line
  • 1 of 5 founding seats
Start via Telegram →

How to buy: click any tier above → the Telegram bot guides you through USDC.e payment on Polygon mainnet (chain ID 137)your X-API-Key is minted and DM'd to you automatically, usually within seconds — use it as the X-API-Key header on GET /api/v1/signals. No email, no account. Want to try before paying? The keyless free tier returns the top 10 of /api/v1/signals with full provenance on every signal.

Proof suite

ResourceWhat it gives you
verify.htmlRecompute the SHA-256 of any signal in your browser. If the digest matches the published hash, the observation (signal) cryptographically existed at committed_at_utc.
iran-timeline.htmlLive example: OIN's Iran observations sealed in Bitcoin block 952139 on 2026-06-03, seven days before the war-resumption headline. Every line falsifiable.
scorecard.jsonMachine-readable accuracy scorecard. Auto-recomputed daily on on-chain ground truth, with Wilson LCB published. Accuracy is not statistically distinguishable from the base rate — no directional edge claimed. No cherry-picking.
signal_commits.jsonlFull append-only commit log. Download and audit locally. One JSON commit per line.

Free API tier

No key required: GET https://oracleintel.io/api/v1/signals returns the top 10 free signals with SHA-256 provenance on each. Add an X-API-Key header from a paid tier to unlock the full stream.

curl https://oracleintel.io/api/v1/signals
# → top 10 signals, no key needed

curl -H "X-API-Key: <your-key>" https://oracleintel.io/api/v1/signals
# → full stream, paid tier

Analytics & research only — not financial, legal, or sanctions advice. No performance guarantee, no win-rate claim. OIN's directional accuracy is published openly at scorecard.json and is not statistically distinguishable from the base rate — no demonstrated directional edge. Receipts, not returns. OIN tracks public on-chain wallets only.