Find the crypto perpetuals whose price has spent the most time at its current level — on Binance USDⓂ Futures.
For every token the scanner pulls the last 1500 candles on the 1h, 4h and 1d timeframes and counts how many of those candles had the token's current price between their high and low. The three counts are summed into a single stability score. The higher the score, the longer price has lingered around where it trades today.
Two ways to run it: a clean Streamlit dashboard or a plain console script.
- Live token universe — pulls every tradable USDⓂ Futures symbol straight from the exchange, filtered by USDT or USDC.
- Pick exactly what you scan — multiselect with Select all / Clear, so you never waste API calls on tokens you don't care about.
- Ranked, readable results — sorted highest-first with an inline stability bar, summary metrics and one-click CSV export.
- Live progress — per-token progress bar while the scan runs.
- One cross-platform launcher — a single script detects your OS (Windows / macOS / Linux), installs dependencies, opens the dashboard in your browser, and a matching stopper shuts everything down.
- Console mode — the original behaviour is preserved for scripting.
USDT scan — ranked results with stability bars
USDC scan — switch quote currency in one click
git clone https://github.com/AlexBraguta/stability.git
cd stabilityCopy the template and fill in your keys:
cp .env.example .envBINANCE_API_KEY=your_api_key_here
BINANCE_API_SECRET=your_api_secret_hereThe scanner reads only public market data (klines, ticker, exchange info), so it runs even with placeholder keys. Real keys simply raise your rate limits.
| OS | Start | Stop |
|---|---|---|
| Windows | double-click launch.bat |
double-click stop.bat |
| macOS / Linux | ./launch.sh |
./stop.sh |
| Any OS | python launch.py |
python stop.py |
launch.py installs the dependencies, starts the server and opens a
browser tab automatically. When you're done, run the stopper — it
cleanly kills the server (no orphan processes).
On macOS/Linux, make the wrappers executable once:
chmod +x launch.sh stop.shPrefer plain output? Scan the built-in token list and print a sorted table:
python main.pyFor a token at current price P:
score = (# of last 1500 1d candles where low < P < high)
+ (# of last 1500 4h candles where low < P < high)
+ (# of last 1500 1h candles where low < P < high)
A high score means price keeps returning to (or never left) this region — a sign of a strong, well-tested level. A low score means the current price is unusual relative to recent history.
stability/
├── app.py # Streamlit dashboard
├── main.py # Console entry point + built-in token list
├── stability.py # Core logic (shared by both entry points)
├── config.py # Loads API keys from .env
├── launch.py # Cross-platform launcher (OS-aware)
├── stop.py # Cross-platform stopper
├── launch.bat/.sh # Double-click / shell wrappers
├── stop.bat/.sh # Double-click / shell wrappers
├── requirements.txt
├── .env.example
└── results/ # Historical console-run dumps
- Python 3.9+
- Packages (installed automatically by the launcher):
streamlit,pandas,numpy,binance-futures-connector,python-dotenv
- Each token costs 4 API calls (3 × 1500 candles + 1 ticker). Scanning the full ~570-symbol USDT universe is slow and rate-limit sensitive — pick a subset in the dashboard for quick runs.
- Market is USDⓂ Futures only (perpetuals), not spot.
.envand the runtime PID file are git-ignored — your keys never leave your machine.
This project is for informational and educational analysis only. It is not trading or financial advice, not an investment recommendation, and not a solicitation to buy or sell any asset. The stability score is a statistical observation of past price data and says nothing about future performance. Markets are risky and you can lose money. You are solely responsible for any decisions you make — always do your own research. The software is provided "as is", with no warranty of any kind.
Released under the MIT License — free to use, copy, modify and distribute.