Canli Validation: Check whether a backtest is real before you trust it

Hi Product Hunt, I'm Arhan. I build Canli Capital, an open quant research project that publishes every number with the file behind it, including the strategies that failed. The most common way to fool yourself in trading is to try many ideas and keep the best one; its Sharpe ratio looks great because you picked it. Canli Validation corrects for that: tell it how many things you tried and it tells you how much of the result is luck. It runs as an MCP server, so you can ask Claude or Cursor "is this backtest real?", or as a plain API with a free key. Results come with a signed receipt anyone can recompute. I tested the corrections themselves on a zoo of synthetic searches where the truth is known, and published where they break (https://canlicapital.com/researc...). I'd love to hear where it's wrong.