# Qanat

> MIT MCP for quant factor processing and backtesting

**Category:** AI Agents  
**Pricing:** open-source  
**URL:** https://github.com/fidetolabs/qanat  
**GitHub:** https://github.com/fidetolabs/qanat  
**Added:** 2026-10-04

---

## The Hook

Qanat is a MIT data platform for quant factors — YAML in, process/store/replay history, serve results to your agent over MCP so research stays askable instead of notebook-only.

## What's Good

MIT (fidetolabs/qanat). PyPI `qanat-fdtl` v0.3.0 (HTTP 200). `uv tool install qanat-fdtl` then `qanat init --demo` builds synthetic strategies with no network/key; `qanat mcp` for Claude Code and other clients; optional Streamable HTTP. Aimed at daily/weekly rebalance horizons. ★~287.

## The Catch

Beta quant niche — not a general agent framework. Your real market data sources may need keys/network even though Qanat itself makes no calls. Vs generic data MCP / notebook stacks: catch is factor pipeline + backtest priced after fees, exposed only through MCP. Pin PyPI; demo data is synthetic.

## Verdict

MIT quant factor MCP when agents should operate a backtest store — not a general RAG/data MCP; bring your own market sources beyond demo.