You've got an idea for a trade. Maybe it's "buy when the trend turns up on gold" or "fade the spike after London opens". The problem is the gap between that idea and something a computer can actually run. Traditionally that gap was code. You either learned to program, or you paid someone who could, or you gave up.
You don't have to do any of those things anymore. Here's how the no-code workflow actually works, one step at a time.
Step one: get clear on what you want before you touch anything
This is the part most people skip, and it's the part that matters most. Software can search through thousands of strategy variations a minute. It can't read your mind. So you have to tell it what "good" looks like to you.
Think about a few things. Which market do you want to trade? One pair is plenty to start, say EURUSD or gold. What timeframe suits your life? If you can't watch charts all day, an hourly or daily strategy makes more sense than a one-minute scalper. How much pain can you sit through? A strategy with long flat patches is a different animal to one that grinds away constantly, even if both end up in roughly the same place.
Write these down as constraints, not hopes. "I trade gold on the hourly. I don't want it risking more than a small slice of the account on any one trade. I want something I can leave running while I'm at work." That's a brief. Vague goals produce vague results, no matter how clever the tool.
Step two: let the software generate and test, not you
Here's where the no-code part earns its keep. Instead of you hand-building rules and checking them one at a time, the software does the searching. You give it your market, your timeframe and your constraints, and it generates a pile of candidate strategies, then tests every one against historical price data.
That testing is the whole point. A backtest runs a strategy over past data to see how it would have behaved. On its own that's not enough, because anything can look good if you only show it the data it was built on. The trick a decent tool uses is keeping some history hidden during the build, then testing on that unseen slice afterwards. A strategy that holds up on data it never saw is more interesting than one that only shines on the data it was tuned to. That's the difference between a real pattern and a fluke.
This is also where grading comes in. Nebula scores each strategy 0 to 100 on how robust it looks in testing. Read that carefully. It's a measure of how well the thing held together when stress was applied. It isn't a forecast of profit and it isn't a promise about the future. A high score means the pattern survived poking. It doesn't mean it'll make money next month. Nothing can tell you that, and anyone who claims otherwise is selling something.
Step three: look at the results like a sceptic
When the software hands you candidates, resist the urge to grab the prettiest equity curve. Ask boring questions. How many trades did it actually take? Ten winning trades is noise. A few hundred is a sample worth thinking about. Did it hold up on the hidden data or only the data it was built on? Does the behaviour match your brief, or did you ask for steady and get something that swings wildly?
The goal here isn't to find a perfect strategy. There isn't one. The goal is to throw out the ones that are obviously fragile and keep a small handful that look like they're doing something genuine. You're filtering, not hunting for a winner.
Step four: export to MetaTrader and watch before you commit
Once you've got a candidate you trust, you turn it into something a trading platform can run. Nebula exports an Expert Advisor, which is just MetaTrader 5's name for an automated strategy file. You drop it onto a chart and the platform runs the rules for you.
Do not skip the demo step. Run the EA on a demo account first and watch how it behaves with live prices and real spreads, plus the small frictions a backtest can gloss over. This is also the stage worth knowing about if you're eyeing a prop firm. Most firms allow MT5 EAs, but the rules differ from firm to firm. The challenge fees and the capital are at risk, and passing is never guaranteed. It comes down to you and the conditions, not the software. A tool can build and test a strategy. It can't make any strategy more likely to pass an evaluation.
Where to start
The honest way in is to try the workflow on a single pair before you spend anything. Define a simple brief, generate some candidates, see how the testing and grading work, and get a feel for what robust actually looks like next to lucky. You'll learn more from one afternoon of doing it than from a week of reading about it.
Nebula's free download opens straight onto a single-pair backtester. You don't need an account, a card or any code to start. Download Nebula free and test your first strategy.