Edge Scanner

A real-time intraday stock scanner for US equities that runs on your own machine. It streams 1-minute bars for thousands of symbols, checks the setups you define on every bar, and shows the alerts in a browser dashboard you lay out yourself. For day traders who want their own scans instead of someone else's alert list.

• Your machine, your keys. No account, no subscription, no telemetry. Market data comes from your own Alpaca or Charles Schwab login, and everything it builds stays in data/. Any other provider can be added behind the same interface.
• Setups without code. Compose them in the dashboard from about 45 triggers and a list of conditions. Changes apply on the next bar, with no restart.
• It tells you why. For any stock, Setup Check shows what each setup did in the last few minutes and exactly which filter stopped an alert.
• A dashboard you arrange. Free-floating windows, several saved screens, linked symbols, and a WebSocket feed other programs can subscribe to.
• Free and MIT licensed. No paid tier.

Quick start · Watch the walkthrough · User guide · Releases · Privacy

What this is not: not financial advice, not a signal service, and not a broker. It never places an order. The sample setups are starting points, not a strategy. The screenshots below come from a replayed past session, so the alerts are synthetic in the sense that they were produced by a replay: they are not anyone's real account, positions or orders.

I build my own stock scanner setups, free and open source: why I stopped paying for a scanner, how the setups are built, and the app running on a live session.

Privacy and your data

Edge Scanner is local-first. The scanner and the dashboard listen on localhost only, and there is no login because nothing is exposed to the network. Requests must also be addressed to this machine by name, and settings can only be changed from a page this machine served, so a web page you visit cannot reach the scanner through your browser.

• Stays on your machine: your API keys in .env, the setups, universe filters, watchlists and screens you build, the bar caches, and the alert archives. All of it under data/, which is gitignored.
• Goes out, to your own accounts: market data requests to Alpaca or Charles Schwab, signed with your keys. They see the symbols you scan, as any market-data client would.
• Goes out, unauthenticated: company fundamentals from Yahoo Finance, and per-symbol news headlines from Yahoo Finance and Nasdaq RSS. Those services see the symbol being requested and your IP. Turn the news ones off with NEWS_RSS_SOURCES= in .env.
• Goes out, once at startup: one anonymous request to GitHub's public releases page, to tell you in the dashboard when a newer release exists. GitHub sees your IP and the scanner version, nothing else. Nothing is downloaded or installed; the update is git pull. Turn it off with UPDATE_CHECK=0 in .env.
• Never collected: no analytics, no crash reports, no usage data. The project has no server.

What's inside

• Custom setups, no code. A catalog of about 45 triggers (candle patterns, level breaks, crosses, VWAP and EMA behaviour, opening range, momentum, relative strength against SPY), each with its own options and parameters. Combine them with AND, OR or "at least N of", add conditions the stock must meet, and pick the universe the setup runs on.
• Universe filters. Named screens over price, liquidity, ATR, float, sector, relative volume and more, deciding which stocks a setup may alert on.
• Dashboard windows. Alert tables, charts, rankings (gainers, losers, most active, pre-market lists, new highs and lows), news, stock info, watchlists, Setup Check and a market clock.
• One data connection. Everything runs in one process on one market-data websocket: Alpaca by default, or Charles Schwab, which is free with a brokerage account. Providers sit behind one DataFeed interface, so anything that can stream 1-minute bars and answer for history can be plugged in: see adding a data provider.
• A feed for other programs. ws://localhost:7777/ws/alerts, with server-side filters by source, setup or symbol.
• Extensible. Optional engine plugins can add built-in setups and data providers (scanner/plugins.py).

Screenshots

Price Action: alerts on the left, the one you clicked pinned above the table and marked on the intraday chart, a daily chart and news on the right, and a second link group along the bottom where rankings drive their own chart.

Pre-Market: pre-market gainers, losers and volume leaders. Click any symbol and the chart, the daily chart, news and stock info all follow it.

Config: every setup in one list, with how it fires, when it may fire, and the universe filter it runs on. Alerts, parameters, a plain-English summary and a per-stock check are the other tabs.

Requirements

• Python 3.11 or newer, and Node.js 20 or newer to build the dashboard
• Market data, either:
  • an Alpaca account (paper is fine). The default ALPACA_FEED=sip needs their paid market-data plan. The free ALPACA_FEED=iex works, but it is one exchange, so volume and relative volume read far lower and volume-based setups fire much less; or
  • a Charles Schwab brokerage account, free, using DATA_PROVIDER=schwab. It covers the whole universe: real 1-minute bars for the 300 most liquid symbols (Schwab's limit) and bars built from Schwab's quotes for the rest, which are close but not identical. See the user guide for the one-time login, which Schwab expires every 7 days.

Those two ship because they are the two the project runs on and tests against. Another provider is one class and one line in a registry: adding a data provider.

Tested on Windows. macOS and Linux should work through start_scanner.sh.

Quick start

On Windows, clone the repository and run setup.bat. It checks your Python and Node versions, installs everything, builds the dashboard and creates .env. Then add your keys to .env and run start_scanner.bat.

By hand, or on macOS and Linux:

git clone https://github.com/simonro/edge-scanner.git
cd edge-scanner
python -m venv .venv

Activate the environment (.venv\Scripts\activate on Windows, source .venv/bin/activate elsewhere), then:

pip install -r requirements.txt
cp .env.example .env

Edit .env and set ALPACA_API_KEY, ALPACA_SECRET_KEY and ALPACA_FEED, or set DATA_PROVIDER=schwab and the Schwab keys.

Build the dashboard once:

npm --prefix dashboard-v2 install
npm --prefix dashboard-v2 run build

Run it (start_scanner.bat on Windows):

./start_scanner.sh

Open http://localhost:7777.

Every install starts empty. No watchlists, no alert history and none of anyone else's setups. You get two sample screens (Pre-Market and Price Action) so the dashboard is not a blank page. The first start downloads a year of daily bars and 20 days of 5-minute bars for the whole universe, which takes 10 to 20 minutes; later starts use the cache in data/.

Optional, to start from working examples instead of an empty setup list:

python scripts/install_setup_library.py

The full guide is USER_GUIDE.md.

Updating to a new release

git pull
pip install -r requirements.txt
npm --prefix dashboard-v2 install
npm --prefix dashboard-v2 run build

Your .env and everything in data/ (setups, universe filters, screens, watchlists, caches, alert archives) are left alone. Release notes call out anything that needs a one-time step.

The dashboard shows a small "vX.Y.Z available" link next to the feed status when a newer release exists. To be told by GitHub instead, use Watch, then Custom, then Releases on the repo page.

Configuration

.env keys, all optional unless you use the feature:

Key	Default	What it does
ALPACA_API_KEY, ALPACA_SECRET_KEY	none	Alpaca market data and Benzinga news
ALPACA_FEED	sip	sip (paid, consolidated tape) or iex (free, one exchange)
DATA_PROVIDER	alpaca	alpaca, schwab, or a provider you add yourself
SCHWAB_APP_KEY, SCHWAB_APP_SECRET, SCHWAB_CALLBACK_URL	none	Charles Schwab market data
NEWS_RSS_SOURCES	yahoo,nasdaq	Free per-symbol news feeds. Empty turns them off

Layout

Path	What
scanner/	the engine: data providers, per-symbol state, indicators, setups, universe filters, API
scanner/trigger_catalog.py	every trigger a custom setup can use
scanner/conditions.py	the conditions setups and universe filters can check
scanner/data/	market-data providers (alpaca.py, schwab.py) behind one interface
scanner/custom_setups_defaults.json	sample setups, seeded on first run
scanner/setup_library.json	the setup library (scripts/install_setup_library.py)
scanner/plugins.py	extension points for optional engine plugins
scripts/run_live.py	the live scanner
dashboard-v2/	the React dashboard
data/	runtime data: caches, settings, your setups, alert archives (gitignored)

Development

python -m pytest
npm --prefix dashboard-v2 run lint
npm --prefix dashboard-v2 run dev

npm run dev serves the dashboard with hot reload on http://localhost:5174 and proxies the API to a running scanner on 7777.

Contributing

Bug reports, new triggers and pull requests are welcome. CONTRIBUTING.md covers the setup, the tests a change needs, and how to add a trigger, a condition or a data provider. Please report security problems privately, as described in SECURITY.md, not in a public issue.

Who made this

A day trader who wanted scans that match how he actually trades, and a dashboard that fits one ultrawide monitor. Built by describing the problems to Claude Code.

• YouTube: @tapetoedge
• X: @tapetoedge

Educational software, not financial advice. No affiliate relationship with Alpaca, Schwab or anything else shown here. You are responsible for your own trading decisions, and the software comes with no warranty.

License

MIT, see LICENSE. Third-party notices, including the TradingView Lightweight Charts attribution, are in NOTICE.