Quant Finance MCP Server

HPSILab Quant Finance MCP Server

Stock analysis and options analytics for AI agents

Connect HPSILab to Claude, Cursor, ChatGPT workflows, or your own agent. Get prediction probability, Monte Carlo range, IV regime, options pressure, and backtest context in one structured call.

Start with the NVDA demo, then connect the MCP server when you are ready to test it inside your own agent.

MCP + RESTAgent-ready JSONOptions pressureGlobal coverage: US, Japan, Hong Kong, China A-shares, SingaporeInformational use only
PartnerHalcyon Waters

Options volatility

Today's IV Regime Rankings

Updated daily

IV regimes help separate price momentum from market-implied risk before reading any AI signal.

Public IV regime snapshot is waiting for the next backend refresh.

Final layer

Today's AI Signals

Updated daily

Percent is the model's estimated next-session up probability: 55%+ Bullish, 45% or lower Bearish, otherwise Neutral.

Public AI signal snapshot is waiting for the next backend refresh.
What Agents Get

One quant finance engine, multiple decision lenses.

Generic feeds tell an AI agent what happened. HPSILab gives it probability, options structure, simulation range, and model evidence so the answer is easier to verify.

Market Coverage

Options data coverage by market

Options-derived signals (IV, gamma, flow) depend on whether the exchange lists individual-stock options. Price and momentum tools are available everywhere HPSILab supports a ticker, and prices are shown in each market's local currency based on the ticker suffix.

MarketTicker suffixCurrencyOptions dataNotes
United StatesUSD · $Full coverageOptions chains, implied volatility, and flow analysis.
Japan (JPX).TJPY · JP¥AvailableIndividual-stock options exist; liquidity varies by name.
Hong Kong (HKEX).HKHKD · HK$AvailableIndividual-stock options exist; liquidity varies by name.
China A-shares.SS / .SZCNY · CN¥Not availablePrice/momentum analysis only — no listed single-stock options market exists for A-shares.
Singapore (SGX).SISGD · S$Not availablePrice/momentum analysis only — SGX lists structured warrants and single-stock futures instead of stock options.
MCP Tool Surface

Built for stock research agents, not raw data dumps.

analyze_stockFree
Aggregates every quant tool into one JSON stock analysis: direction signal, direction score, bullish/bearish factors, plain-English summary.
get_ai_predictionFree
Next-day probability the stock closes up, a buy/watch/sell-lean signal, and how strongly the underlying models agree.
get_iv_radarFree
Implied-volatility structure: how expensive options are, volatility squeezes, call/put skew.
get_monte_carloFree
Monte Carlo price simulation for the next ~10 trading days: likely range and odds of finishing higher.
get_option_pressureFree
Option-chain pressure map for the nearest expiry: Max Pain, dealer Gamma Wall, likely weekly high, extreme squeeze target.
get_equity_curvesFree
Backtest performance of the quant strategy across a watchlist: Sharpe ratio, max drawdown, win rate, total return.
generate_stock_imagesFree
Generates stock-report chart images (PNG) and returns their URLs.
Pro
get_pretrade_risk_scanPro
Full pre-trade risk scan for a stock before entering a position.
generate_stock_research_reportPro
Presentation-ready markdown research report combining analysis and charts for a stock.
Workflow

From vague market questions to structured answers.

01

Connect once

Add the HPSILab MCP endpoint to Claude, Cursor, ChatGPT tooling, or your internal research agent.

02

Ask a market question

The agent requests structured quant context instead of guessing from headlines and price charts.

03

Return defensible context

HPSILab responds with probability, options analytics, simulation range, and risk language the agent can cite.

Trust Boundary

Built for research, not blind trading.

Informational use only

HPSILab returns research context for agents and analysts, not buy/sell instructions.

Structured outputs

Responses are JSON-ready for agent workflows, dashboards, alerts, and review logs.

Review before action

Agents can draft watchlists and research reports, while humans remain in the loop.

Install

Give your AI agent a market research layer.

Use MCP for agents and REST for your own backend, scripts, dashboards, and alert pipelines. Developer unlocks API keys, higher limits, and production integration support.

Install
pip install hpsilab-mcp
PyPI versionSmithery
Quick start
from hpsilab_mcp import HpsiMcpClient client = HpsiMcpClient() client.analyze_stock("NVDA")
Get started

Sign in

HPSILab | Quant Finance MCP for AI Stock & Options Analytics