Labsco
MCP SERVER

FinanceKit MCP

by vdalhambra

Quotes, indicators, option chains and portfolio risk for stocks and crypto — the self-hosted server needs no account and no API key.

Equity & Fund Market DataVerified
Summary
The analysis layer, not just the data feed.

Most free market APIs hand back raw series and leave the maths to you; here the indicators, risk ratios and correlations are computed server-side, so a question like whether a position is overbought or how badly two holdings move together is one call rather than a spreadsheet. Two limits worth knowing before you lean on it: quotes come from Yahoo Finance through yfinance rather than an exchange feed, and responses are cached, so a 60-second-old quote is by design.

What it is

A market data and analysis server built on Yahoo Finance and CoinGecko, with the indicators calculated locally rather than fetched. It covers a single ticker, a basket, or a whole portfolio, and answers with numbers plus a plain-English read of what they say.

What you get
  • Quotes for one ticker or a batch, with change, volume, 52-week range, P/E and market cap, plus a company's sector, industry, revenue and margins
  • Crypto by CoinGecko id: price, market cap, 1h/24h/7d change, all-time-high data and supply, the top coins by market cap, today's trending coins, and a search that finds the id from a name or symbol
  • Technical analysis on a ticker — RSI, MACD, Bollinger Bands, SMA 20/50/200, EMA 12/26 — with Golden Cross, Death Cross and overbought/oversold flagged rather than left to be read off
  • OHLCV history over a chosen period and interval, with min, max, average price and total volume summarised
  • Assets compared side by side on return and volatility over the same window
  • Portfolio analysis from SYMBOL:SHARES pairs: current value, weight distribution, sector breakdown and per-position performance
  • Risk metrics for a symbol against a benchmark — 95% Value-at-Risk, annualised Sharpe and Sortino, Beta and maximum drawdown
  • A correlation matrix of pairwise Pearson correlations across a basket, which is the number diversification actually turns on
  • A market snapshot — S&P 500, NASDAQ, Dow, Russell 2000 and the VIX with the day's biggest movers — and the 11 GICS sectors ranked by return through their SPDR ETFs
  • Earnings dates with EPS estimate against actual for recent quarters, and the option chain with strikes, volume, open interest, implied volatility and Greeks where available
Requirements

Nothing to sign up for when you run it yourself: no API key for either stocks or crypto. Python 3.11 or newer, started over stdio as uvx --from financekit-mcp financekit, or pip install financekit-mcp and run financekit. Stock data comes from Yahoo Finance through yfinance and crypto from CoinGecko's free tier, with responses cached — quotes 60s, crypto 2 minutes, history 1 hour, company info 24 hours. A hosted endpoint exists as an alternative to running it, metered by calls per month with a free tier of 100. MIT.

Setup effort

One command — uvx --from financekit-mcp financekit