Built and killed a lot of strategies using claude. These are the 2 I'm actually running on MNQ right now
Been building and throwing away strategies on MNQ for a while. Most die in testing. I tried retest entries, divergence entries, basic buy and sell one contract hold, New york ORB, and some others, I have all pine scripts saved. Ive taken about 2 automated payouts. one for 500 and one for 1000. Fully automated fully passed the account and taken payout(2-3 weeks). But then also blew my accounts right after. These two survived and are the only ones I have wired up live right now, so here's the whole thing including the numbers I don't love. Im trying to have a running strategy that can hold multiple funded accounts without risking blowing them.
The pipeline
TradingView (Pine v5) -> alert with {{strategy.order.alert_message}} -> TradersPost webhook -> prop account.
A third alert forks to a Google Apps Script -> Sheet -> my own journaling app so every fill is logged independent of the broker. This journaling file is more so a TradingHub idea I had also thought up. Having a dedicated area of space for prop firm trading, with a news calender, whatever the average trader would need to be omnipresent with the market as well as add any of their sites, not needing to scroll through many tabs. that was the idea. Still building upon this and would love some insight.
**1. EMA Pullback Structure Strategy
MNQ 5m, long and short. Enters once at full size, exits in three sized legs: a quick TP, a runner, and a bigger target. Force-flat at 12:50 PT so nothing is held overnight.
- Stik MR(Stochastic Mean Reversion)
MNQ daily, long only, mean reversion. Scales into as many as 4 units on successively deeper dips. Holds overnight and over weekends, which is exactly why it can never run on a funded prop account - Topstep force-flats at 3:10pm CT and an overnight position is an instant violation. It runs on a dead account as a signal target only. Would need a live account with 5-10K drawdown to safely run. This scales up though, compounding interest.
There is one issue I ran into. I could only run backtesting on tradingview 5min for 7 weeks back. So I built a backtesting engine with Claude, primarily for pine scripts. Has some issues here and there with running some concepts but its easily fixable.
Me and some friends also set up a little discord for a community dedicated for prop firm algo trading help. We also set up a discord News Bot that delivers real time important news dedicated for MNQ,MGC,ES and other futures indexes.
This post is simply a reach out and share of my work so far. Would love to hear feedback and speak on this concept further! If you wanna interact with all the things in this post join this discord or just reply here.