Trading dashboard and research stack

One program you run on your computer or server: open it in the browser and use it as a trading assistant for workflow and research. The same process serves the dashboard and the JSON API behind it—no separate app pair to coordinate for daily use.

Scanner (what’s live now)

While markets are relevant to your setup, the scanner watches your saved symbols, applies technical and rules-based logic, and surfaces candidates with scores, checks, and context—so you can see what looked interesting and why, alongside levels and indicators when alerts exist. New passes can be kicked off from the UI or driven by your scheduler configuration.

Quant / research (historical honesty)

Backtesting, walk-forward optimization (WFO), and supporting statistics live alongside the scanner in quant/. That lane answers questions like “did this idea hold up over history?”—a different cadence than intraday scans. Methodology and limits are documented under docs/ (start with docs/STRATEGY_AND_EDGE.md); product direction and phases are summarized in docs/timeline.md.

This stack is educational tooling: it helps you organize scans and validate ideas systematically; it does not replace your judgment, risk controls, or compliance obligations.

---

Python application with a FastAPI server that serves a web dashboard (site/) and JSON APIs under /api/*.

scanner_core/ runs the market scan pipeline (regimes, indicators, scoring). quant/ holds backtesting, walk-forward optimization, statistics, and data helpers. backend/app.py wires routes, static files, and API endpoints. tests/ contains pytest suites; scripts/ has pool validation and multiple-testing utilities.

Quick start

cd c:\Users\__\___\___
pip install -r requirements.txt
copy .env.example .env
# optional: FINNHUB_API_KEY, ALPHA_VANTAGE_API_KEY, POLYGON_API_KEY, SUPABASE_*

python -m uvicorn backend.app:app --host 127.0.0.1 --port 8000

Open http://127.0.0.1:8000/ — same origin serves the UI and APIs (/api/health, /api/scan/latest, POST /api/scan/run, POST /api/quant/backtest, /api/quant/tutorial).

Or run run_web.ps1 from the repo folder (starts the same server).

Browser shows error -102 or “connection refused”

Nothing is listening on port 8000. Start the server first, keep that terminal open, then reload the page.

Optional checks

curl.exe http://127.0.0.1:8000/api/health

Expect JSON with "status":"ok". If embedded previews inside some tools fail but curl works, use Chrome or Edge at http://127.0.0.1:8000/.

Heavy walk-forward batches (multi-day runs): run_wfo_batch.ps1 → python -m quant.run_wfo ....

Layout

Path	Role
backend/app.py	FastAPI app, static mount, REST routes
scanner_core/	Scan pipeline, regimes, indicators
quant/	Backtest / WFO / stats / data helpers
site/	Dashboard HTML/CSS/JS
scripts/	Pool validation, multiple-testing helpers
tests/	pytest

Strategy methodology and edge notes live in docs/STRATEGY_AND_EDGE.md.