A stock screening platform with multi-methodology scans across US, Hong Kong, India, Japan, Korea, Taiwan, mainland China A-share, Germany, Canada, Singapore, Malaysia, and Australia markets, AI-assisted research, theme discovery from social and news feeds, and real-time market breadth analysis. The supported deployment path is a single-tenant server stack built around Docker, PostgreSQL, Redis, and nginx.

Static demo: https://xang1234.github.io/stock-screener/

The static page is for demo purposes only. It is a read-only daily snapshot with reduced functionality compared with the full application, which includes live workflows such as chatbot, themes pipeline, watchlists, and the full interactive backend.

Scan and track twelve markets:

• 🇺🇸 United States — NYSE, NASDAQ, AMEX, S&P 500
• 🇭🇰 Hong Kong — HSI
• 🇮🇳 India — NSE, BSE
• 🇯🇵 Japan — Nikkei 225
• 🇰🇷 Korea — KOSPI, KOSDAQ
• 🇹🇼 Taiwan — TAIEX
• 🇨🇳 Mainland China A-shares — SSE, SZSE, BJSE
• 🇩🇪 Germany — XETRA, DAX
• 🇨🇦 Canada — TSX, TSXV
• 🇸🇬 Singapore — SGX
• 🇲🇾 Malaysia — Bursa Malaysia (Main + ACE), FBM KLCI
• 🇦🇺 Australia — ASX, S&P/ASX 200

Each market runs on its own exchange calendar (XNYS / XHKG / XNSE / XTKS / XKRX / XTAI / XSHG / XETR / XTSE / XSES / XKLS / XASX) with independent Celery refresh queues and locks, so US, Asia-Pacific, and Europe can hydrate in parallel without stepping on each other. Switch markets from the scan control bar; mixed-universe results are tagged with per-row colored badges.

Market picker in the scan control bar — pick US, HK, IN, JP, KR, TW, CN, DE, CA, SG, MY, or AU and scope to an exchange or index

Color-coded per-market badges in the Symbol column — US (blue), HK (green), JP (yellow); Taiwan, India, Korea, China, Germany, Canada, Singapore, Malaysia, and Australia follow the same pattern

Deep-dive: ASIA v2 ADRs & runbooks

Run Minervini, CANSLIM, IPO, Volume Breakthrough, Setup Engine, and Custom scans simultaneously with composite scoring across 80+ configurable filters. Save filter presets and export results to CSV.

Results table with composite scores, RS sparklines, multi-screener ratings, and per-row classification columns — GICS Sector, IBD Industry, market themes, and group rank

StockBee-style advance/decline analysis with SPY overlay, daily movers (stocks up/down 4%+), and multi-period trend visualization across quarterly, monthly, and 34-day windows.

Breadth chart with SPY price overlay and daily movers

197 industry groups ranked by relative strength with top movers identification (1W/1M/3M/6M), historical rank charts, and constituent stock analysis.

Industry group rankings with movers panel

A MarketSmith/Bloomberg-style quadrant view of the same 197-group dataset: every industry group (or GICS-sector roll-up) is plotted by RS-Ratio vs RS-Momentum, with a weekly tail and direction arrow tracing its path through Leading → Weakening → Lagging → Improving. One screen answers "what's rotating in, what's rolling over." Filter by name or current rank to focus on the groups you care about, and click any dot to drill into its constituents.

RRG: industry-group rotation with direction-arrowed weekly tails

Visual performance tracking with RS and price sparklines (30-day trends), price change bars across 7 time periods, drag-and-drop organization with folders, and full-screen chart navigation.

Watchlist with sparklines and price change visualization

Groq-powered research chat with optional Tavily/Serper web search, persistent conversation history, and tool-augmented investigation.

AI chatbot with conversation sidebar and research tools

AI-powered market theme identification from RSS, Twitter/X, and news feeds. Tracks trending vs. emerging themes, monitors constituent stocks, and alerts on momentum shifts.

Theme discovery with rankings and emerging themes panel

cp .env.docker.example .env.docker
# Edit .env.docker:
#   BACKEND_IMAGE=ghcr.io/<owner>/stockscreenclaude-backend
#   FRONTEND_IMAGE=ghcr.io/<owner>/stockscreenclaude-frontend
#   APP_IMAGE_TAG=v1.2.0
#   SERVER_AUTH_PASSWORD=choose-a-long-random-password
#   GROQ_API_KEY=...
ENABLED_MARKETS=US,HK,CN scripts/docker-compose-enabled-markets.sh --env-file .env.docker -f docker-compose.yml -f docker-compose.prod.yml -f docker-compose.release.yml pull
ENABLED_MARKETS=US,HK,CN scripts/docker-compose-enabled-markets.sh --env-file .env.docker -f docker-compose.yml -f docker-compose.prod.yml -f docker-compose.release.yml up -d --no-build
# Open http://localhost

This path deploys the tagged v1.2.0 GHCR images instead of building locally. After the stack is up, the UI opens to a first-run bootstrap screen — see First-Run Bootstrap for the staged pipeline and market selection.

For local development or contributor laptops, use the default local compose stack instead:

cp .env.docker.example .env
# Edit .env: set SERVER_AUTH_PASSWORD and add at least one LLM API key (e.g., GROQ_API_KEY)
scripts/docker-compose-enabled-markets.sh up

Full guide with homelab, VPS, and GHCR deployment options: Docker Deployment

The default Compose file defines worker services for every supported market, but market-specific workers are behind Compose profiles. Use the helper script to start only the workers for the markets in ENABLED_MARKETS:

ENABLED_MARKETS=US,HK,CN scripts/docker-compose-enabled-markets.sh up -d

For ENABLED_MARKETS=US,HK,CN, Docker starts US/HK/CN market job and user scan workers. IN/JP/KR/TW/DE/CA/SG/MY/AU worker containers are not created. The global data-fetch worker listens only to data_fetch_shared,data_fetch_us,data_fetch_hk,data_fetch_cn.

The first-run bootstrap wizard still persists runtime choices in Postgres. Keep the wizard's enabled markets within the deployment ENABLED_MARKETS set. To add a market later, update ENABLED_MARKETS and recreate the stack:

ENABLED_MARKETS=US,HK,CN,TW scripts/docker-compose-enabled-markets.sh up -d

See the Development Guide for full backend + frontend + Celery setup.

On first launch the app boots into a setup screen — no pre-seeded database required. Pick a primary market (the one that opens first) and any additional enabled markets to hydrate in the background, then click Start bootstrap.

Bootstrap performance note: selecting multiple enabled markets starts separate universe, price, fundamentals, breadth, group-rank, and scan work for each market. This can noticeably slow first load and ongoing data updates on smaller hosts or when upstream market-data providers throttle requests. For the fastest first run, start with one primary market and enable additional markets after the workspace is ready.

Primary-market dropdown and enabled-markets checkboxes on first launch

The orchestrator runs a staged Celery pipeline for the primary market:

1. Universe refresh — seeds the market's symbol list (S&P 500 / Russell / NDX for US via refresh_stock_universe; official exchange feeds for HK / IN / JP / KR / TW / CN / CA / DE / SG / MY / AU via refresh_official_market_universe).
2. Benchmark + price refresh — imports the GitHub daily price bundle first, accepts recent stale bundles during bootstrap, then live-fetches only missing/current-session gaps (7d top-up for stale symbols, 2y only for no-history symbols).
3. Fundamentals refresh — quarterly and annual financials.
4. Breadth calculation — StockBee-style advance/decline with gap-fill.
5. Group rankings — IBD-style relative strength across 197 industry groups.
6. Feature snapshot (US only) — daily feature rollup used by the Setup Engine.
7. Initial autoscan — seeds a first scan with the default profile so you land on populated results.

Per-stage progress with per-market queue status while the pipeline is running

The workspace opens as soon as the primary market reaches ready. Secondary markets keep hydrating in the background on their own queues (data_fetch_{us,hk,in,jp,kr,tw,cn,de,ca,sg,my,au}) so you can start scanning immediately. Scans against a market that's still refreshing return HTTP 409 market_refresh_active, and scans against a market whose data is not current yet return the usual stale-data response instead of triggering a first-run block.

State is persisted in AppSetting under runtime.primary_market, runtime.enabled_markets, and runtime.bootstrap_state (not_started → running → ready). To re-run the wizard, reset runtime.bootstrap_state to not_started.

The AI chatbot requires at least one LLM provider API key. Scanning and all other features work without any keys.

Optional web search keys (TAVILY_API_KEY, SERPER_API_KEY) enable the chatbot's research mode.

Full reference: Environment Variables

• 12 supported markets — US, Hong Kong, India, Japan, Korea, Taiwan, mainland China, Germany, Canada, Singapore, Malaysia, Australia — with per-market exchange calendars, independent refresh queues, and scan-time freshness guards
• First-run bootstrap wizard with live staged progress and background hydration of secondary markets
• 6 screening methodologies with composite scoring (Minervini, CANSLIM, IPO, Volume Breakthrough, Setup Engine, Custom)
• 80+ configurable filters with saved presets across fundamental, technical, and rating categories
• AI chatbot with Groq-first routing, web search research mode, and persistent conversations
• Theme discovery from RSS, Twitter/X, and news sources with AI clustering and lifecycle tracking
• Market breadth dashboard with StockBee-style indicators and historical trends
• 197 IBD industry groups ranked by relative strength with movers and constituent analysis
• Watchlists with RS/price sparklines, multi-period change bars, and drag-and-drop organization
• MCP integration for AI copilot workflows with 12 tools via stdio and Streamable HTTP (details)
• TradingView-style charts with candlestick OHLC and technical overlays
• CSV export for scan results
• Dark and light mode UI
• Docker deployment with PostgreSQL, auto-HTTPS, and GHCR image releases

Backend: FastAPI, SQLAlchemy, Alembic, Celery, Redis, PostgreSQL Frontend: React 18, Vite, Material-UI, TanStack Query / Table, Recharts Data: yfinance, Finviz, Alpha Vantage, SEC EDGAR, official X API (optional)

This software is for educational and research purposes only. It is not financial advice. Always do your own research and consult with a licensed financial advisor before making investment decisions.

Released under the Apache License 2.0.