Infrastructure
Market intelligence an agent can read.
Point a Model Context Protocol client at one URL, sign in with a Data Plan, and prices, volatility surfaces and the instrument catalogue arrive as tools a model can call directly. Access requires no API key or SDK, and the model reads the same data as every other client.
Every call is authorised against your own account, so an agent only ever reads what your plan already covers. Retrieval comes with an active plan; running backtests needs the Strategy Backtest subscription.
Ask, in plain English
Fetch the latest 7-day vs 30-day volatility smile and compare the results.
Maps to
data_retrieval
{ "frequency": "1h", "params": { "request_type": "iv", "base_assets": ["BTC"], "exchanges": ["blockscholes"], "models": ["SVI"], "level": { "iv_type": "money", "values": [0.85, 0.9, 0.95, 1, 1.05, 1.1, 1.15] }, "tenors": ["7d", "30d"] }, "time_range": { "range_type": "latest" } }
Response · 7D vs 30D
ts 1786543200
- 0.8548.7144.55-4.16
- 0.9041.1939.51-1.68
- 0.9533.3135.41+2.10
- 1.0027.6533.20+5.55
- 1.0529.8532.97+3.12
- 1.1035.1833.89-1.29
- 1.1540.3935.30-5.09
Block Scholes composite · SVI · shared 7-point grid · Sample data (past capture)
Illustrative typing, real schema and real values — the same call, and the same 7D/30D smile, published across this site.
From a sentence to data
Ask a question. Get calibrated data back in seconds.
- /01
Ask
Questions are asked in plain English, naming an asset, a tenor or a comparison.
- /02
The agent picks a tool
It reads our tool list, chooses the right one and fills in the asset, tenors and model itself.
- /03
Block Scholes answers
Calibrated values are returned from the same pipeline that serves the API and the Terminal.
- /04
You get the answer
Charted or explained in seconds, and every number traces back to the call that fetched it.
What an agent can readVolatility surfaces →Market data →Indices →Historical data →
The tool list
Three tools, all read-only.
Prices and volatility
data_retrieval
Query spot, perpetual, futures, and options price data. Retrieve implied volatility surfaces, mark prices, funding rates, open interest, volumes, and more across multiple exchanges.
What does the BTC volatility term structure look like right now?
index.iv · 6 tenors
- 7D31.02
- 14D33.84
- 30D38.31
- 90D43.55
- 180D48.05
- 365D51.82
index.iv · BTC BSIV term structure
Sample data (past capture)The catalogue
get_instruments_tool
Look up available instruments per exchange, including listed expiries, strikes, and option types.
What BTC option strikes are currently listed?
get_instruments_tool
41 strikes, spanning $52,000 to $78,000, this expiry.
get_instruments_tool · strike ladder
Strategies, run
backtest_strategy
Execute strategy backtests with trade-by-trade PnL, hedge attribution, and funding rate incorporation. Supports spot, perpetual, futures, and options positions with theoretical or listed pricing modes.
Requires the Strategy Backtest subscription.
Backtest a BTC 30-day straddle at listed bid/ask.
backtest_strategy · illustrative
- Model mid0.0421 BTC
- Listed bid/ask0.0435 BTC
- Edge−3.2%
backtest_strategy · equity
Illustrative
Declared on every tool
read-onlyidempotentnon-destructive
Connecting
One URL, authenticated over OAuth.
The whole integration is the block on the right. Every connection authenticates over OAuth — the first time a client connects, it redirects you to sign in to your Block Scholes account, and the tools it can reach follow from what that account already entitles. Access is managed through the account, so there is no key to mint, rotate or store in an agent’s environment.
- Claude
- Web and Desktop · recommended
- Cursor
- Editor-native
- Any MCP client
- Standard protocol
{
"mcpServers": {
"blockscholes": {
"url": "https://mcp.blockscholes.com/mcp"
}
}
}OAuth · sign in on first connect
What you get
One call reaches the full query range.
The three tools give access to the range below. A single data_retrieval call names the asset, the venues, the model and the level required: implied volatility as an index, by delta, by strike or by moneyness, as skew or butterfly, or the underlying SVI parameters, at minute or hourly granularity, for the latest print or a specified window.
Anatomy of one call
data_retrieval(…)
Asset types
asset_type
- spot
- perpetual
- future
- option
Price the ETH June future
Volatility levels
level
- index
- delta
- strike
- moneyness
- skew
- butterfly
- SVI params
BTC 25-delta skew at 30 days
Granularity
frequency · range
- minute
- hourly
- latest
- range
- named points
Hourly ATM vol over the last week
Models
models
- SVI
- spline
The same smile, fitted with a spline
backtest_strategy is the same engine behind the Strategy Backtester — see that page for how a multi-leg strategy is described, priced and attributed.
Delivered over
MCP
Use cases
What teams build on AI / MCP
Get started
Integrate AI / MCP into your own systems.