A cleaner front door to your exported research.
Each card below links to a bundle-level dashboard index. The page scans the current `reports/*/dashboards/index.html` layout and summarizes the strongest candidate from each results bundle.
EURUSD=X • 1y • 1d
SPY • 1y • 1d
GLD • 1y • 1d
AAPL • 1y • 1d
AMZN • 1y • 1d
SOL-USD • 1y • 1d
ETH-USD • 1y • 1d
From market data to ranked dashboards in four moves.
The system is built for high-throughput research: generate ideas, reject weak fits quickly, then inspect only the strongest candidates in a human-readable format.
Choose data
Set ticker, history, interval, fees, and stop logic in the GUI or script.
Generate candidates
Run random search or grid search over indicator-based entry and exit combinations.
Validate and rank
Score candidates with weekly metrics, drawdown rules, and optional three-phase validation.
Review exports
Open the dashboard index, compare the top ideas, and inspect exported Python strategy files.
Everything needed for users, reviewers, and maintainers.
The repo now includes a layered documentation set: a quick README, a complete user guide, and a deeper code manual that explains architecture, extensions, scoring, and outputs.
Fast onboarding
Short setup notes, file map, outputs, and the right next links for a first-time visitor.
How to use the app well
Covers settings, workflows, result interpretation, fallback dashboards, and troubleshooting.
How the system works inside
Documents architecture, rule parsing, backtesting, ranking, validation, exports, and safe extension points.