Skip to content

dcimring/cryptobee-dashboard

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Cryptobee

Stockbee-style momentum dashboard for the crypto market — sibling of the stockbee project, sharing its architecture (Python + SQLite + FastAPI + one static HTML page) but rebuilt for a 24/7, volatility-scaled asset class.

How it differs from the stock version

Stockbee Cryptobee
Universe ~5,000 US common stocks (Tiingo directory) Top ~150 coins by market cap (CoinGecko top-500 ∩ Massive USD-pair coverage), stablecoins/wrapped excluded
Burst definition fixed 4% daily move move beyond the coin's own 20-day ATR% (breadth 1×, scan entry 1.5×)
Episodic pivot 8% on 2× volume, earnings auto-flagged 3.5× ATR% on 2× dollar volume, catalyst manual
Liquidity floor 100k shares $5M dollar volume ($10M for EPs)
Calendar Tue–Sat, weekend/holiday guards, split re-backfills every day, no adjustments (no splits/dividends)
Benchmark SPY BTC
Theme rollup Nasdaq industry CoinGecko category (Meme, DeFi, AI, …)
Extra gauge BTC dominance (% of universe market cap)

All thresholds live in pipeline/config.py.

Setup

cp .env.example .env   # add POLYGON_API_KEY (Massive.com free tier)
uv sync
uv run python -m pipeline.universe --categories   # universe + category tags
uv run python -m pipeline.ingest_daily --days 400 # one-time backfill (~90 min, rate-limited)
uv run python -m pipeline.compute                 # breadth + scans
uv run uvicorn app.main:app --port 8001           # dashboard at localhost:8001

Nightly automation: scripts/nightly.sh via launchd (com.cryptobee.nightly, daily — crypto has no weekends).

Pipeline

universe (CoinGecko top-N + BTC dominance snapshot, weekly category tags) → ingest_daily (Massive.com grouped crypto bars, one call per UTC day) → compute (breadth, momentum_burst / anticipation / episodic_pivot scans) → FastAPI (app/main.py) → static dashboard (app/static/index.html).

Regime thresholds start as the stock dashboard's tuned values; re-tune against forward BTC returns with uv run python -m pipeline.backtest_regime --grid once a few hundred days of breadth history exist.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages