Weather High-Temp Sniper

Automated trader for Polymarket weather highest temperature markets. Uses Simmer SDK's importable endpoint for market discovery and trades only during local 9-10 AM window when YES price exceeds threshold. Supports Telegram notifications.

Automated trading bot for Polymarket weather markets focusing on "highest temperature" predictions. Uses Simmer SDK's modern market discovery workflow to find and import new weather markets, then executes trades during a specific time window.

Features

• Market Discovery: Uses /markets/importable endpoint to find new Polymarket weather markets
• Duplicate Prevention: Checks /markets/check before importing to avoid duplicates
• Active Market Fetching: Retrieves active weather markets via get_markets(import_source='polymarket', tags=['weather'])
• Time-Window Trading: Only trades during local 9-10 AM based on market location's timezone
• Fallback Window: Optional 10:00-10:05 fallback if configured
• Telegram Notifications: Real-time alerts for trades, errors, and daily reports
• Configurable Parameters: All trading parameters tunable via environment or automaton

How It Works

1. Market Discovery & Import

The bot runs a discovery phase every 6 hours to find new weather markets:

1. Fetches daily import quota from Simmer
2. Queries list_importable_markets(venue="polymarket", q="highest temperature")
3. Filters markets using is_weather_market_question() heuristics
4. For each candidate, checks if it already exists via check_market_exists()
5. Imports new markets using import_market() until quota is exhausted

2. Active Market Scanning

After discovery, the bot fetches active weather markets:

python

markets = client.get_markets(
    import_source="polymarket",
    status="active",
    max_hours_to_resolution=48
)

It then locally filters to only weather-related markets using the same heuristics.

3. Trading Logic

For each weather market:

1. Extract location from the question string and detect timezone (via CITY_TZ_MAP)
2. Check time window: Only trade if local time is 9:00-9:59 AM (or fallback 10:00-10:05)
3. Get current price (YES probability)
4. Check threshold: Price must be ≥ ENTRY_PRICE_THRESHOLD
5. Validate position size: Cost (shares × price) must be ≤ MAX_POSITION_USD
6. Execute trade: place_order(market_id, "buy", shares, price, slippage)
7. Record state: Mark market as traded today to avoid re-entry

4. Position Management

• No position in the same market on the same day
• All trades tracked in sniper.state.json
• State persists across restarts

Configuration

All parameters are configurable via environment variables or via the OpenClaw automaton (--set flag):

Parameter	Environment Variable	Default	Description
entry_price_threshold	SIMMER_SNIPER2_ENTRY_PRICE_THRESHOLD	0.35 (35%)	Minimum YES price to trigger a buy
max_position_usd	SIMMER_SNIPER2_MAX_POSITION_USD	2.50	Maximum USD risk per trade
shares_per_order	SIMMER_SNIPER2_SHARES	5	Number of YES shares per order (min 5 on Polymarket)
slippage_tolerance	SIMMER_SNIPER2_SLIPPAGE	0.15 (15%)	Maximum acceptable slippage
scan_interval_seconds	SIMMER_SNIPER2_SCAN_INTERVAL	300 (5 min)	How often to scan for opportunities
fallback_at_local_10	SIMMER_SNIPER2_FALLBACK_10AM	true	Enable 10:00-10:05 fallback window
telegram_enabled	SIMMER_SNIPER2_TELEGRAM_ENABLED	false	Enable Telegram notifications
telegram_chat_id	SIMMER_SNIPER2_TELEGRAM_CHAT_ID	""	Your Telegram chat ID
report_interval_seconds	SIMMER_SNIPER2_REPORT_INTERVAL	240 (4 min)	How often to send periodic reports

Setting Configuration

Via environment variable:

bash

export SIMMER_SNIPER2_ENTRY_PRICE_THRESHOLD=0.40
export SIMMER_SNIPER2_MAX_POSITION_USD=5.00

Via command line (overrides config.json):

bash

python sniper.py --set entry_price_threshold=0.40
python sniper.py --set max_position_usd=5.00

Via OpenClaw automaton (persistent tuning): The skill integrates with the Simmer Automaton — configuration changes via automaton are automatically persisted and synced.

Installation & Setup

Prerequisites

• Python 3.8+
• Simmer API key (from simmer.markets/dashboard)
• Polymarket linked wallet (bot will auto-link on first run)
• Telegram bot token and chat ID (optional, for notifications)

Install Dependencies

bash

cd polymarket-weather-high-temp-sniper-1
pip install -r requirements.txt

Configure Environment

Create .env file in the skill folder:

bash

# Required
SIMMER_API_KEY=your_simmer_api_key_here

# Optional (Telegram)
TELEGRAM_BOT_TOKEN=123456:ABC-DEF1234ghIkl-zyx57W2v1u123ew11
TELEGRAM_CHAT_ID=123456789  # Your Telegram user ID
TELEGRAM_ENABLED=true

Or set environment variables directly in your shell/automaton config.

Test Telegram

bash

python sniper.py --test-telegram

This verifies bot token and sends a test message.

Usage

Dry Run (Default)

See what would be traded without executing:

bash

python scripts/sniper.py

Live Trading

Execute real trades:

bash

python scripts/sniper.py --live

The daemon will:

• Run discovery phase every 6 hours
• Scan every 5 minutes (configurable via scan_interval_seconds)
• Send Telegram notifications on trades and errors
• Print periodic reports every 4 minutes

Show Current Positions

bash

python scripts/sniper.py --positions

Show Active Configuration

bash

python scripts/sniper.py --config

Adjust Parameters On-The-Fly

bash

python scripts/sniper.py --set entry_price_threshold=0.40
python scripts/sniper.py --set max_position_usd=5.00
python scripts/sniper.py --set scan_interval_seconds=180

Stop the Bot

Press Ctrl+C in the terminal where it's running. A shutdown notification will be sent via Telegram if enabled.

State Persistence

The skill stores its runtime state in sniper.state.json:

json

{
  "date": "2025-01-15",
  "traded_today": ["market_id_1", "market_id_2"],
  "last_report_ts": 1705312345,
  "last_discovery_ts": 1705312000
}

• traded_today: List of markets already traded today (reset at midnight UTC)
• last_report_ts: Timestamp of last periodic report
• last_discovery_ts: Timestamp of last market discovery phase

Important Notes

Timezone Detection

The bot auto-detects timezone from city names in the market question. It recognizes major cities worldwide (see CITY_TZ_MAP in the code). If a city is not recognized, the market is skipped.

Price Source

The bot reads current_probability or external_price_yes from the Simmer market object. Prices are expressed as probabilities (0.0 - 1.0).

Market Question Filtering

Only markets meeting these criteria are considered:

• Must contain "temperature", "temp", "°c", or "°f"
• Must contain "highest" or "high"
• Must NOT contain "lowest", "minimum", "low", "below", "fall below"
• Must NOT contain noise keywords: crypto, bitcoin, election, president, sports, etc.

Import Quota

Simmer imposes daily import limits (typically ~10/day). The bot respects remaining quota and will skip discovery when exhausted.

Position Sizing

Each trade costs: shares_per_order × price. The bot checks this ≤ max_position_usd before buying. Example: With max_position_usd=2.50, shares=5, and price 0.35, cost = 5 × 0.35 = $1.75 ✅.

Minimum Shares

Polymarket requires a minimum of 5 shares per order. The default shares_per_order=5 satisfies this constraint.

Troubleshooting

"SIMMER_API_KEY not set"

Set your API key in .env or environment. Get it from simmer.markets/dashboard → SDK tab.

"Telegram: bot token invalid"

Test with --test-telegram. Verify:

• TELEGRAM_BOT_TOKEN is correct (format: 123456:ABC-DEF...)
• Bot has been started (send /start to your bot in Telegram)
• TELEGRAM_CHAT_ID is your numeric user ID (not username)

Markets Not Being Traded

• Check if current time is within 9-10 AM local time for the market location
• Verify price is ≥ entry_price_threshold (default 35%)
• Ensure cost (shares × price) ≤ max_position_usd
• Check logs for skipped markets and reasons

Quota Exhausted

The bot will log when daily import quota (10/day) is exhausted. Wait for the next UTC day or request quota increase from Simmer.

No Positions Showing

Use --positions to fetch current positions. Ensure wallet is linked (bot attempts auto-link on first run).

Support

For issues or feature requests, refer to the OpenClaw documentation or contact the skill author.