Compact research environment for exploring risk-managed trading strategies across equities and crypto. The focus is on answering three questions:

1. Which strategy + parameters perform best per asset when risk controls are enforced?
2. How do those results behave when visualised at scale (outliers, distributions, summary stats)?
3. What actually happened inside a run (entries, exits, portfolio heat) for the top performers?

Everything in the repo exists to serve those flows and provides a clean base for the future portfolio engine.

The scripts write to cache/, visualizations/, and the repo root as needed. Everything is single-asset today; the PortfolioEngine stub anchors future blended portfolio work.

micromamba env create -f environment-simple.yml -y

# Run optimizer + full report pipeline
micromamba run -n trading-bot-simple python main.py --mode report --opt-symbols all --report-top-k 3

# Or run steps separately:
micromamba run -n trading-bot-simple python main.py --mode optimize --opt-mode multi-symbol --opt-symbols crypto
micromamba run -n trading-bot-simple python main.py --mode report-only --report-top-k 5

# Standalone visualization from existing JSON
micromamba run -n trading-bot-simple python main.py --mode visualize

• Cached market data lives in data_cache/. Clear it with python main.py --clear-data-cache.
• Optimiser and tester caches live under cache/ with per-symbol JSON dumps.
• Charts drop into visualizations/ (ignored by git).
• End-to-end lab reports land in reports/<timestamp>/ with JSON, Markdown, and charts.

.
├── main.py                     # Thin CLI wrapper (optimize / visualize / inspect / portfolio)
├── src/trading_bot/            # Core package
│   ├── data.py                 # Yahoo Finance ingestion + caching
│   ├── multi_asset_tester.py   # Strategy sweeps per symbol (uses risk-managed strategies)
│   ├── optimizer.py            # Parameter optimiser built on the tester
│   ├── results_visualizer.py   # Summary charts (boxplots, outlier handling)
│   ├── top_performer_visualization.py  # Trade + risk timeline plots
│   ├── risk_management.py      # Risk manager, heat/drawdown monitors
│   ├── risk_managed_strategy.py / risk_managed_strategies.py
│   ├── portfolio_engine.py     # Portfolio scaffold for future work
│   └── ...
├── tests/                      # Pytest suite (invoked via `pytest` or `run_tests.py`)
├── docs/                       # Focused notes tied to the new layout (see below)
└── scripts                     # Utility entry points (generate fixtures, legacy smoke tests)

The old long-form guides have been retired. The remaining references are:

• docs/overview.md – high-level architecture diagram + quick reminders
• docs/portfolio-roadmap.md – notes for the upcoming blended portfolio engine
• CHANGELOG.md – user-facing changes

micromamba run -n trading-bot-simple pytest -q              # full suite
micromamba run -n trading-bot-simple pytest tests/test_optimizer.py -q
python run_tests.py --module test_risk_management

Pytest is configured with pythonpath = src, so package imports (trading_bot.*) work from anywhere.

• Flesh out PortfolioEngine with cross-asset allocation, shared risk budgets, and rebalancing logic.
• Promote optimiser outputs into a simple API (e.g. trading_bot.optimize.run(...)) for notebooks.
• Persist analytics (summary tables, charts) to a structured reports/ directory with timestamps.
• Add PostgreSQL or DuckDB sinks for long-term result analysis.

Feel free to keep the repo lean—any file that doesn’t serve the three workflows gets pruned.