LIVE · 9 exchanges · multi-asset · agent-first

library 10,352 products · research leads 644 · history 14y

01 / LORE — The Agent Data Layer

Your AI
isn't dumb.
Your data is.

Most crypto data was built for dashboards, analysts, and human traders. LORE rebuilds it for agents — structured, labeled, explained, and ready to use.

Get access → See it work no free tier · cancel anytime · from $75/mo

Exchanges

9

USD-M perps

150

Spot assets

65+

History

2012→now

Built for

agents

LIVE TAPE

BTC-PERP · BINANCE

Last

94,127.5

24h

+2.74%

live · 9 venues normalized agent-ready

Products

10,352

Edge maps

748

Leads

644

agent-first · MCP-ready · public catalog · cancel anytime

" scroll — the receipts are real "

02 / The Missing Layer for AI Agents — Sight · Memory · Judgment · Direction

Raw crypto data
gives your AI facts.
LORE gives it intelligence.

Most crypto data products stop at raw delivery — candles, trades, and APIs built for humans. LORE adds the missing layer: structured market context, historical memory, validation, and clear starting paths so your AI can reason, not just ingest.

NORMALIZED HISTORICAL VALIDATED AGENT-READY

THE LORE TRANSFORM

input → intelligence → output

i. RAW INPUT

other APIs

# what other crypto APIs deliver
{
  "ts":    "2024-03-12T14:22Z",
  "open":  71240.5,
  "high":  71890.0,
  "low":   71180.0,
  "close": 71840.5,
  "vol":   2841.2
}

YOUR AI

Invents indicators. Hallucinates patterns. Overfits one chart.

↓ →

ii. LORE INTELLIGENCE

ACTIVE · tx 12,847,329

◎ Sight

chart_state · regime · tradability

⟲ Memory

historical_analogs · pattern_replay

⊠ Judgment

matched_controls · base_rate · source_trust

➤ Direction

paths · bundles · agent_workflows

PROCESS

Structures, labels, validates, and connects raw data into something your AI can reason over.

↓ →

iii. AGENT OUTPUT

with LORE

YOUR AI'S RESPONSE

"Found 12 BTC breakouts in trending_low_vol regimes, beating their matched control by +1.8%.

18 historical analogs ready.

tradability: good
source_trust: high"

YOUR AI

Does research. Cites evidence. Knows when to stop.

your AI struggles the missing layer → your AI does research

03 / The Tape — Every print, level, and liquidation

Markets move in sequences.
Your AI should see the sequence.

LORE aligns trades, books, funding, liquidations, open interest, and positioning into one timestamped market feed. So your agent can study cause, pressure, and reaction — not just final prices.

TRADES ORDER BOOK FUNDING LIQUIDATIONS OPEN INTEREST POSITIONING BASIS CROSS-VENUE

PRINT TAPE

BTC-PERP | live

TIME SIDE PRICE SIZE

Δ 1m +0.18% / 1,247 prints

L2 BOOK

top 8 | live

SPREAD $0.50 94,127.25

Imbalance +0.34 ASKS

LIQUIDATIONS

cluster · 5m

LONGS

-$8.2M

142 wallets

SHORTS

-$1.8M

38 wallets

Largest 24h $2.4M @ 92,840 LONG

i.

Sequenced.

Replay the market exactly as it happened, event by event, across venues.

ii.

Normalized.

One schema for trades, books, funding, liquidations, open interest, basis, and positioning.

iii.

Validated.

Gap checks, stale-feed warnings, sparsity flags, and source reliability scores built in.

04 / Setup — Paste a prompt. You're done.

Paste this into Codex,
Claude, or Cursor.
You're done.

No SDK to learn. No docs to read. Your AI handles the install and writes the queries — you just ask questions.

LORE ships an MCP server, a Python SDK, and a TypeScript SDK. Pick whichever your AI is wired into. The prompt below works with all of them.

PROMPT · COPY ME

# paste anywhere your AI is listening

Add LORE to this project. Install the SDK, configure my
API key from LORE_API_KEY, and run one sample query:
find BTC setups that beat their matched control.
Show me the result.▮

REST + SDK

Get an API key →

pip install lore

MODEL CONTEXT PROTOCOL

Add as MCP server →

claude mcp add lore

step 1.

Your AI reads the catalog.

10,352 dataset cards. 53 product families. Plain-English explanations of what every dataset is and how to use it safely.

step 2.

Your AI writes the query.

Joins the right products. Avoids leakage. Respects timeframe alignment. Filters on regime, source trust, sample size.

step 3.

Your AI gets a real answer.

Not "I see BTC went up." A ranked list of setups, regime context, base-rate comparison, similar historical analogs.

05 / Who LORE is for — Bot makers · Researchers · Founders

LORE is for

Start with what you’re building. LORE pulls the right crypto data, labels, context, and AI-ready memory from one marketplace — so your agent can stop parsing and start thinking.

⌬

BUILD A BOT

Build the ultimate trading bot.

For solo bot builders, indie quants, vibe coders shipping a working bot.

YOU GET

·7 bot bundles (momentum, breakout, mean rev, funding, scalping, rotation, cognition)
·Full historical depth (150 perps, 2012 → today)
·Regime context, cross-asset signals, tradability
·Pattern Memory Replay — every time it happened before
·644 ranked Research Lead Cards as starter ideas

try this prompt:
"Build me a momentum bot that only trades BTC when funding flips negative AND OI is expanding. Use LORE for the data."

Indie · $75/mo See bundles →

⌖

AI RESEARCH

Research a quant-level strategy.

For curious traders, AI agents, researchers using AI to find and study setups.

YOU GET

·Edge Atlas v7 — 748 maps, 8M+ episode rows
·644 Research Lead Cards (prompt-ready dossiers)
·Pattern Memory Replay — replay any setup
·Matched Control Atlas — base-rate comparison
·Research Evidence Graph — your AI's compass

try this prompt:
"Show me ranked BTC setups from 2024–2026 that beat their matched control AND had high source trust. Group by regime."

Indie · $75/mo Open the stack →

⊞

BUILD A PRODUCT

Create the next great crypto startup.

For indie SaaS founders, Discord/Telegram operators, dashboard builders, copilot makers.

YOU GET

·Full API + commercial use rights
·Webhook delivery for live signals
·Higher rate limits to power your app
·3 seats (Startup) or 25 seats (Pro)
·Up to ~250 (Startup) or ~10K (Pro) end users

try this prompt:
"I'm building a Discord bot that posts BTC setups to my channel. Use LORE to pull high-priority research leads every hour."

Startup · $599/mo Build with LORE →

06 / The Agent Workflow — A real research session

Ask one question.
LORE turns it into a research workflow.

LORE turns one question into a full research process. Your AI can search the market, study similar moments from the past, check whether the pattern actually worked, and come back with an answer you can use.

SESSION_2026-05-08 · BTC_BREAKOUT_RESEARCH · AGENT_REPLAY

REPLAY · 4 TURNS · 6.3 SEC

01

14:22:01.0

01 / ASK

▶ YOU

"Find BTC breakouts that beat their base rate."

◉ LORE resolved in 412ms

await lore.research_leads({
  asset:            "BTC",
  setup_type:       "breakout",
  compare_to:       "base_rate",
  include_controls: true,
})

EDGE ATLAS MATCHED CONTROLS

→ 12 candidates found · avg excess +1.8%

◇ AGENT synthesizing

Found 12 BTC breakout setups in low-vol regimes beating their base rate by +1.8%. Strongest candidate has 18 historical analogs across 2020–2024. Want me to investigate it?

02

14:22:03.1

02 / DRILL DOWN

▶ YOU

"Yes — show me the strongest one's history."

◉ LORE resolved in 587ms

await lore.pattern_replay({
  setup_id:               "btc_breakout_low_vol_03",
  include_analogs:        true,
  include_regime_context: true,
})

PATTERN REPLAY REGIME CONTEXT

→ 18 analogs replayed · 14 passed quality gates

◇ AGENT synthesizing

The strongest pattern appeared 18 times. 14 passed source-quality checks. Median forward return was +3.4%, with the best outcomes occurring when open interest expanded without funding overheating.

03

14:22:05.2

03 / VALIDATE

▶ YOU

"Is it real, or just the market already going up?"

◉ LORE resolved in 331ms

await lore.matched_controls({
  setup_id:     "btc_breakout_low_vol_03",
  control_type: "same_regime_no_breakout",
  horizon:      "7d",
})

MATCHED CONTROL ATLAS SOURCE TRUST

→ 173 controls checked · excess return +1.8% · source trust high

◇ AGENT synthesizing

Matched controls show the setup outperformed similar low-vol periods without the breakout by +1.8%. Tradability score: good. Source trust: high. Leakage risk: low.

04

14:22:07.3

04 / SYNTHESIZE

▶ YOU

"Turn it into a research note I can test."

◉ LORE resolved in 1.42s

await lore.lead_dossier({
  setup_id:         "btc_breakout_low_vol_03",
  include_prompt:   true,
  include_graph:    true,
  include_controls: true,
})

LEAD CARDS EVIDENCE GRAPH AGENT WORKBENCH

→ 1 dossier created · 644 evidence nodes traversed · leakage-safe

◇ AGENT complete

Dossier complete: 18 analogs, regime context, tradability rating, matched-control delta of +1.8%, source reliability flags, and a prompt-ready summary. Ready for your backtester.

TOTAL TIME

6.3 sec

AGENT TURNS

4

LORE PRODUCTS USED

9

FINAL OUTPUT

Backtester-ready dossier

That’s the difference between an API response and an agent workflow. LORE gives your AI the layers to search, replay, validate, and synthesize — until the output is ready to test.

07 / Vs. the Incumbents — What changes when the customer is your AI

They sell crypto data.
We sell what your AI can use.

Kaiko, CoinAPI, Tardis, Coinglass — built for human traders looking at dashboards. Here’s what changes when the customer is your AI agent.

QUESTION YOUR AI KEEPS ASKING

KAIKO · COINAPI · TARDIS

LORE

"What regime is the market in?"

✕

Regime Context Layer

546 pre-labeled products

"Does my AI see what a trader sees?"

✕

Chart Cognition Layer

3,637 products · 119M chart-state rows

"Has this setup happened before?"

✕

Edge Atlas v7

8.3M outcome rows · 748 strategy maps

"Did it actually beat its base rate?"

✕

Matched Control Atlas

644 leads · 173 beating controls

"Which dataset should I even start with?"

✕ Read 200 pages of docs

Agent Navigation

4 tiers · 6 start-here workflows

"Can I afford this on an indie budget?"

✕ $2,000–$6,000/mo

$75/mo Indie

full API + full data

08 / The Catalog — What's actually in the library

What's in the library.
Eight categories. All built for AI.

53 product families across 10,352 products — every category structured and labeled for AI agents, from the raw market sequence up through ranked research dossiers.

Products

10,352

Families

53

Coverage

2012→

LAYER 1 ▦

Market Activity

Real-time and historical record of every trade, price level, and order book event across 9 venues.

· Print tape (trades)
· Level-2 order books
· Kraken BTC true-L3 queue lifecycle
· Mark, index, premium
· Native venue timestamps

81.7M perp bars · L2 + L3

LAYER 1 ◈

Derivatives

Funding, open interest, basis, and forced-close events for the perp universe.

· Funding rate history
· Funding & carry feature pack
· Open interest (OI)
· Liquidation & deleveraging (live forward feeds)
· Perp / spot basis

813K funding · 82.1M mark/index

LAYER 1 ⇕

Positioning & Flow

OI changes, crowding, buy/sell aggression, absorption, large-trade activity.

· Long / short ratios
· OI delta + crowding
· Trapped longs / shorts
· Buy / sell aggression
· Absorption · large trades

66.8M position · 118.9M flow

LAYER 2 ⇄

Cross-Exchange

Whether a move is real confirmation across venues or one exchange acting alone.

· Venue confirmation ratio
· Lead / lag strength
· Cross-exchange dispersion
· Listing & launch dynamics
· Isolated-move flags
· Reference + pair gaps

10.6M rows · 9 venues · 74 products

LAYER 2 ◉

Chart Cognition

What a trader sees on the chart — encoded so an AI agent can read it.

· Support / resistance state
· Trend & channel structure
· Volume profile
· Setup labels (22.8M)
· Plain-English captions

3,637 products · 119M chart-state

LAYER 2 ◐

Regimes & Context

Market state tags and cross-asset context — the environment, not just the price.

· Relative Strength Atlas — peer ranks across the universe
· Regime state · cluster IDs · age · transitions
· Trend / chop / vol regime tags
· Tradability Atlas (execution context)
· Breadth · dispersion · liquidity context

546 regime · 8.3M tradability

LAYER 3 ⌬

Edge Atlas + Research

Strategy-discovery map — episodes, outcomes, replays, matched controls, ranked leads.

· Edge Atlas (748 maps)
· Research Lead Cards (644)
· Pattern Memory Replay
· Matched Control Atlas
· Outcome Atlas — for training, never as a feature

8.3M outcomes · 173 beat controls

LAYER 3 ⊞

Trust + Workbench

Reliability scoring, validation gates, and the leakage-safe pipeline that gets you from data to a trustworthy model.

· Source reliability (4 buckets)
· 18 readiness gates
· Column Safety Manifest — every column tagged safe-input vs outcome-label vs warning
· Model Split Manifests — causal train/val/test splits
· Model-Ready Feature Packs — outcomes excluded by construction
· Agent Navigation + Workbench guidance

9 sources scored · 5,520 marts

TIMEFRAMES + HISTORICAL DEPTH

queryable across every category above

9 timeframes — price-derived where source data supports it

for every market in the library

1m 3m 5m 10m 15m 30m 1h 4h 1d

FULL LIBRARY

2012 → today

14 years · bull, bear, crash, recovery

5+ YEARS DEEP

81 perps

enough history for serious backtesting

3+ YEARS DEEP

142 perps

multi-regime coverage built in

✓

18 readiness gates

all passing

✓

48 structural checks

0 failed

✓

0 blocked products

out of 10,352

✓

Source-neutral exports

no proprietary leakage

ASSET COVERAGE — ALL 150 PERPS · 65+ SPOT

preview · top 32 shown below

BTC

●

ETH

●

SOL

●

HYPE

●

BNB

●

XRP

●

DOGE

●

AVAX

●

ADA

●

LINK

●

MATIC

●

LTC

●

TRX

●

NEAR

●

SUI

●

TON

●

ARB

●

OP

●

PEPE

●

WIF

●

BONK

●

JUP

●

PYTH

●

JTO

●

FET

●

RNDR

●

TIA

●

SEI

●

UNI

●

AAVE

●

APE

●

+118

more

all 150 perps live · 14 years of history where source data supports it verify in the public catalog ↗

Want to go deeper?

Browse the full catalog — every category expanded with internal products, coverage notes, and example queries.

Browse the full catalog →

09 / Pricing — Three tiers, full access from day one

Quant level crypto data.
Priced for indie builders.

Most serious crypto data access lives behind $2,000–$6,000/month contracts. LORE gives builders the full API and full data library from day one, starting at $75/mo. Upgrade only when you need more seats, higher limits, or commercial rights.

★ MOST POPULAR

Indie

solo

$75/mo

For solo builders who need the real data stack.

Comparable data access often costs $2,000–$6,000/mo elsewhere.

✓Full API + full data library
✓9 exchanges · 150 perp markets · 65+ assets
✓Edge Atlas, Lead Cards, Pattern Replay, and Matched Controls
✓MCP server, REST API, Python and TypeScript SDKs
✓1 seat
✓Email support
·Personal use only
·Standard rate limits

Get LORE Indie →

Startup

team

$599/mo

For teams turning agents into products.

Commercial rights, higher limits, and delivery built for real users.

✓Everything in Indie
✓Commercial use license
✓Apps with up to 250 end users
✓3 seats
✓Higher rate limits
✓Webhook and forward delivery
✓Email support

Get LORE Startup →

Pro

scale

$2,499/mo

For companies scaling market intelligence products.

Production-grade access for apps, teams, and larger user bases.

✓Everything in Startup
✓Commercial use for apps with up to 10,000 end users
✓25 seats
✓Highest rate limits
✓Priority support
✓Custom symbol expansion on request

Get LORE Pro →

·	Indie	Startup	Pro
Price	$75/mo	$599/mo	$2,499/mo
Full API access	✓	✓	✓
Full data library	✓	✓	✓
Seats	1	3	25
End users	personal only	~250	~10,000
Commercial use	✕	✓	✓
Rate limits	standard	higher	highest
Support	email	email	priority email

Need more than 10,000 end users?

Custom tiers for products at scale. Email info@loremoney.com with what you’re shipping.

Email us →

10 / Get Access

Your AI
isn't dumb.
Now your data isn't either.

9 exchanges. 150 perps. 14 years of history. 644 ranked research leads. All structured, labeled, explained — pre-built for the AI in your stack. From $75/mo.

Get access → See it work

WHY I BUILT THIS

I spent months trying to build crypto bots with Codex and Claude. The AI kept failing — not because it was dumb, but because the data wasn’t built for it. So I built the data layer the AI actually needed.

SIGNAL

              ╔═════════════╗
              ║  ◉  L O R E ║
              ║             ║
              ║  ┌───────┐  ║
              ║  │ ▓▓▓▓▓ │  ║
              ║  │ ▓███▓ │  ║
              ║  │ ▓▓▓▓▓ │  ║
              ║  └───────┘  ║
              ║      │      ║
              ║   ═══╪═══   ║
              ║      │      ║
              ║      ▼      ║
              ║   ╔═════╗   ║
              ║   ║AGENT║   ║
              ║   ║ DATA║   ║
              ║   ╚═════╝   ║
              ╚═════════════╝

library: 10,352 products

tape: live · 9 exchanges

leads: 644 ranked

gates: 18 / 18 passing

$ pip install lore_

Indie price

$75/mo

Incumbent entry

$2K–$6K/mo

9 EXCHANGES · 150 PERPS · 14 YEARS · 644 LEADS · AGENT-FIRST

Your AI isn't dumb. Your data is.

Raw crypto data gives your AI facts. LORE gives it intelligence.

Markets move in sequences. Your AI should see the sequence.

Paste this into Codex, Claude, or Cursor. You're done.

LORE is for researchers.

Build the ultimate trading bot.

Research a quant-level strategy.

Create the next great crypto startup.

Ask one question. LORE turns it into a research workflow.

They sell crypto data. We sell what your AI can use.

What's in the library. Eight categories. All built for AI.

Market Activity

Derivatives

Positioning & Flow

Cross-Exchange

Chart Cognition

Regimes & Context

Edge Atlas + Research

Trust + Workbench

Quant level crypto data. Priced for indie builders.

Indie

Startup

Pro

Your AI isn't dumb. Now your data isn't either.

You found us.

Your AI
isn't dumb.
Your data is.

Raw crypto data
gives your AI facts.
LORE gives it intelligence.

Markets move in sequences.
Your AI should see the sequence.

Paste this into Codex,
Claude, or Cursor.
You're done.

LORE is for

Ask one question.
LORE turns it into a research workflow.

They sell crypto data.
We sell what your AI can use.

What's in the library.
Eight categories. All built for AI.

Quant level crypto data.
Priced for indie builders.

Your AI
isn't dumb.
Now your data isn't either.