Market intelligence a model can call.
An MCP server exposing prices, volatility and indices as tools, so an agent retrieves the same numbers as a trading team, with no bespoke integration per model.
Block Scholes1 min read
- Read-only MCP tools
- 3
- Options venues
- 10+
- Tokens
- 150+
- History since
- 2020
/01 The problem
An agent depends on the data it can reach
Giving a model market data usually means writing an integration per model and per data source, with no guarantee that the numbers it reaches match those the trading team uses.
/02 What you get
The same numbers traders read, as tools a model can call.
- One URL, OAuth sign-in, no API key to manage
- Three tools: retrieve, catalogue, backtest
- Read-only, idempotent, non-destructive
- Entitlement-aware — scoped to what your plan covers
/03 How it works
Technical detail is on the AI / MCP page.
Point a Model Context Protocol client at mcp.blockscholes.com/mcp, sign in with a Data Plan, and three tools — data_retrieval, get_instruments_tool, backtest_strategy — arrive read-only and idempotent. This page covers the use case; AI / MCP sets out the endpoint, the schemas and the client config.
/04 Guardrails
Read-only by design
Every tool is registered read-only and idempotent: an agent can retrieve, catalogue and backtest, and cannot place an order, move a position or write anything back. Every call is authorised against the client’s own account, so an agent reads only what its plan covers.
Retrieval comes with an active Data Plan; running backtests needs the Strategy Backtest subscription.
Questions
What teams ask before they build.
What does an agent need to connect?
One URL and a sign-in with a Data Plan. There is no API key to mint or manage.Which tools are available?
Three: data_retrieval, get_instruments_tool and backtest_strategy, all read-only and idempotent.Can an agent run backtests?
Yes, with the Strategy Backtest subscription.
Read next
The data and tools behind this.
- /01
Infrastructure · AI / MCP
Machine access to pricing, volatility and derivatives analytics over the Model Context Protocol.
- /02
Infrastructure · Strategy Backtester
Position, hedge and lifecycle simulation running on the same data infrastructure.
- /03
Data & Analytics · Volatility
Implied and realised volatility, skew, term structure and calibrated surfaces across strike and maturity.