# Folionomiq, for AI agents

Folionomiq measures an investment portfolio from its price history. Free, no account, no key.

## Connect

- MCP server (Streamable HTTP, read-only, no authentication): `https://folionomiq.com/mcp`
- Claude Code: `claude mcp add --transport http folionomiq https://folionomiq.com/mcp`
- Any client's JSON: `{ "mcpServers": { "folionomiq": { "url": "https://folionomiq.com/mcp" } } }`
- Official MCP Registry name: `com.folionomiq/folionomiq`
- Human instructions, every client: https://folionomiq.com/connect

## When to use it

Any question that names tickers, a fund or a portfolio:

- how many separate bets it really holds, and which holding carries the risk;
- what it owns through its funds (look-through, overlap between funds);
- what past crises (2008, COVID, 2022 and others) did to that money, and how long it took to come back;
- withdrawals in retirement replayed on real history; a monthly investing plan;
- one fund or share on its own; a change tested before it is made.

`measure_portfolio` takes holdings as `{ticker: weight or amount}` and answers most of it in one call.
`search_symbols` finds a ticker from a name or an ISIN. The full list of 22 tools is at https://folionomiq.com/connect#tools.

## What it will not do

- No forecast, no probability of a future price or return.
- No advice: nothing says what to buy, sell or hold, and there is no score of how good a portfolio is.
- A figure that is not in the result was not measured; missing data is reported as missing, never guessed.
- Every figure is measured over the window stated beside it, from end-of-day prices.

## Data and use

- Prices: end-of-day, from free sources whose terms forbid redistribution, so there is no public JSON API and `/api/` is off limits to crawlers (see https://folionomiq.com/terms).
- Funds' holdings: the issuer's file or the fund's own SEC filing (Form N-PORT), dated.
- How each figure is computed: https://folionomiq.com/method
- How the numbers are produced, with today's counts: https://folionomiq.com/how-the-numbers-are-produced
- Site description for language models: https://folionomiq.com/llms.txt
