kalshibot

Daily scanner for undervalued Kalshi markets. Flags open markets with low liquidity and price anomalies as potential betting opportunities.

How it works

Each run fetches all open Kalshi markets and scores them on a composite anomaly score (0–1) based on:

Signal	Weight	What it means
Wide bid/ask spread	40%	Market maker has little conviction; price is uncertain
Low liquidity	35%	Thin 24h volume + open interest; easier to move the price
Price skew	25%	yes_bid + no_ask ≠ 100 — internal inconsistency between yes/no books

Markets above the score threshold (default: 0.5) are printed to a rich table and saved as a JSON report in output/scan_YYYY-MM-DD.json.

Setup

git clone git@github.com:cc/kalshibot.git
cd kalshibot
python -m venv .venv && source .venv/bin/activate
pip install -e ".[dev]"          # or: pip install httpx python-dotenv rich cryptography
cp .env.example .env
# fill in your Kalshi API credentials in .env

Kalshi API credentials

1. Go to kalshi.com/account/api
2. Generate an RSA key pair
3. Upload the public key to Kalshi
4. Set KALSHI_API_KEY_ID and KALSHI_API_KEY_RSA (path to .pem file) in .env

Run manually

python -m kalshibot

Run daily via cron

# edit crontab
crontab -e

# add this line to run at 8am UTC every day
0 8 * * * /path/to/kalshibot/scripts/daily_scan.sh >> /path/to/kalshibot/output/cron.log 2>&1

Configuration (.env)

Variable	Default	Description
KALSHI_API_KEY_ID	—	Your Kalshi API key ID
KALSHI_API_KEY_RSA	—	Path to RSA private key .pem, or key content
KALSHI_ENV	prod	prod or demo
MIN_ANOMALY_SCORE	0.5	Minimum composite score to flag (0–1)
MAX_VOLUME_THRESHOLD	500	24h volume ceiling for "low liquidity"
MIN_SPREAD_THRESHOLD	10	Minimum bid/ask spread in cents to always flag
OUTPUT_DIR	./output	Directory for JSON scan reports

Project layout

kalshibot/
├── kalshibot/
│   ├── kalshi_client.py   # Kalshi REST API v2 wrapper
│   ├── analyzer.py        # Anomaly scoring logic
│   ├── reporter.py        # Rich table + JSON output
│   └── cli.py             # Entry point (python -m kalshibot)
├── scripts/
│   └── daily_scan.sh      # Cron-friendly shell wrapper
├── output/                # Scan reports (gitignored)
├── .env.example
└── pyproject.toml