Enter a strategy's rules in Forge, test them on your own price history, and read the verdict.
Oathra tests a robot's rules, not its file. What that means for a robot you wrote, a robot you bought, and how you check the test is of your robot.
An overfitted backtest fits the past and fails on new prices. The checks Forge runs to tell skill from luck, each in one line.
What MetaTrader's Strategy Tester does, from its own page, and what Forge adds to it. The judging of the result.
MetaTrader 5 can hold back a forward period and test on real ticks, and it runs your actual robot. What it leaves to you is the judging. Forge counts how many settings you tried and marks the result down for it, checks the result across many splits of history, and gives one verdict.
Then you learned it in a test and not on a live account. Its trades can still be logged. There is no test to hold it to until a version survives.
Oathra tests a robot's rules, not its file. It does not run .ex4 or .ex5 files. You enter the robot's entry and exit rules once, in the Strategy Studio or in Python. If you do not know its rules, Oathra still logs it robot by robot, without a verdict.
Import your robot's MetaTrader Strategy Tester report. Forge lines the two up trade by trade and shows where they differ.
No code is needed to start. You build a strategy from conditions you pick in the Strategy Studio, and Python is there when you want it.
You make the same change in your own robot. Forge does not write the robot for you.
Oathra can test the idea: build it in the Strategy Studio with no code and test it. A strategy that survives still has to become a robot on your platform, and Oathra does not write that robot for you.
Forge optimizes the settings on one part of history, then judges them on years it has never seen. It counts how many settings you tried and marks the result down for it.
The test uses your broker's own price history, exported from MetaTrader and uploaded by you. Oathra ships no sample prices. Testing on your own price history needs storage, which the free plan does not include.
Validated is a verdict about the past. It held up on data it had never seen, and it passed the checks that it is not luck. It is not a promise about next month.
Killed means the strategy failed one stage of the test, and the verdict names the stage. A killed result is not an error. It tells you which question the strategy could not answer.
About 15 hours of testing, at the standard test size. Every test shows its credit cost before it starts.
Tests run in Oathra's cloud, on the price history you uploaded. Start a test, close the laptop, come back to the result.
Export a tested strategy to MQL5: Coming soon. No dates promised.
The length of a test depends on the length of the history and the number of settings. Every test shows its credit cost before it starts, and it runs in Oathra's cloud, so you can close the laptop.