weather-agent

Trade Polymarket US temperature markets using NOAA forecasts with dynamic confidence, quality filtering, and smart trade sizing for improved accuracy and ris...

Weather Trading (Enhanced)

CRITICAL: Autonomous Trading Skill

This skill places REAL TRADES with REAL MONEY when enabled.

Key information:

• Default State: autostart:false - Will NOT run automatically
• Autonomous Behavior: Runs every 6 hours when manually enabled
• Financial Risk: Can deploy $30-60/day (max $100/day with quality filter)
• No Per-Trade Review: Trades execute automatically without approval

Installation & Security

Credentials

Required:

• SIMMER_API_KEY - Trading API key from simmer.markets/dashboard

Optional (non-secret configuration):

• 6 SIMMER_WEATHER_* environment variables for trading parameters (see Configuration)

No other credentials needed (no wallet keys, RPC endpoints, or cloud credentials).

Optional dependency: If you install tradejournal, it will log trade details. Not installed by default. Inspect source code before installing.

Installation Method

• Manual installation (no automatic scripts)
• Place files in ~/.openclaw/skills/weather-enhanced/
• Create .env file with API key
• No automatic system writes

Dependencies

python-dotenv>=1.0.0  # Optional

Clean dependency list:

• Uses built-in urllib (no requests library)
• No web3, telegram bot, or other unused dependencies
• Optional: tradejournal for trade logging (commented out in requirements.txt)

Network Endpoints

This skill connects to 3 endpoints:

1. api.weather.gov (NOAA)

   • Purpose: Fetch temperature forecasts
   • Data sent: Latitude/longitude (public coordinates)
   • Data received: Weather forecasts (public data)
   • Authentication: None
   • Code location: weather_trader_enhanced.py line ~250-280

2. nominatim.openstreetmap.org (Geocoding)

   • Purpose: Convert city names to coordinates
   • Data sent: City name string (e.g., "Chicago")
   • Data received: Lat/lon coordinates
   • Authentication: None
   • Code location: weather_trader_enhanced.py line ~200-230

3. api.simmer.markets (Trading)

   • Purpose: Execute trades, read portfolio
   • Data sent: SIMMER_API_KEY (bearer token), trade orders
   • Data received: Trade confirmations, positions, balance
   • Authentication: Bearer token
   • Code location: weather_trader_enhanced.py line ~300-350

Note: If you install the optional tradejournal dependency, it may add endpoints.

Security Guarantees

• No other network connections in code
• API key never logged to disk or console
• API key sent only to api.simmer.markets
• No tracking, analytics, or telemetry
• No data exfiltration
• All network calls visible in source (use grep to verify)

API Key Permissions

Your SIMMER_API_KEY should have:

• Read portfolio (required)
• Read positions (required)
• Place trades (required)
• NO withdrawal permissions
• NO account modification permissions

Create least-privilege key:

1. Visit simmer.markets/dashboard → SDK tab
2. Create new API key
3. Select "Trading Only" if available
4. Never grant withdrawal rights
5. Store in .env file (git-ignored)
6. Rotate key after testing

Configuration Storage

• .env file: Contains SIMMER_API_KEY (never committed)
• config.json: Trading parameters only (no secrets)
• No persistent logs
• All data ephemeral (console output only)

Note: Legacy variables WALLET_PRIVATE_KEY and POLYGON_RPC_URL in .env are unused (web3 removed).

Safety Features

• Default: autostart disabled - Must manually enable
• Quality filter (60%) - Skips low-liquidity markets
• Position limits ($5) - Caps risk per trade
• Trade limits (5/run) - Maximum 5 trades every 6 hours
• Entry threshold (15%) - Only buys undervalued markets
• Exit threshold (45%) - Auto-sells at profit

Before Enabling

DO NOT SKIP THESE STEPS

1. Review Source Code

# Inspect main trading logic
cat weather_trader_enhanced.py | less

# Verify network endpoints (should only find 3)
grep -n "urlopen\|Request\|http" weather_trader_enhanced.py

# Check API key usage (should only send to simmer.markets)
grep -n "api_key\|SIMMER_API_KEY" weather_trader_enhanced.py

# Check optional tradejournal usage
grep -n "tradejournal\|log_trade" weather_trader_enhanced.py

2. Understand Platform Behavior

OpenClaw metadata controls:

• autostart:false = Skill will NOT run on startup (safe default)
• cron:"0 */6 * * *" = Schedule (executes only when autostart enabled)
• Enable by editing line 5: change autostart:false to autostart:true
• Verify this method with OpenClaw docs: https://docs.openclaw.ai/
• Manual file editing prevents accidental activation
• Alternative: Run manually without enabling autostart

When enabled: Runs every 6 hours (12am, 6am, 12pm, 6pm) When disabled: 100% inactive

3. Verify API Key Permissions

Check your Simmer API key:

• Can read portfolio? (required)
• Can place trades? (required)
• Can withdraw funds? (should be NO)
• Can modify account? (should be NO)

If your key has withdrawal permissions, create a new trading-only key.

4. Test Live Trading

# Check balance first
python scripts/status.py

# Execute live trades (start with small balance)
python weather_trader_enhanced.py --live --smart-sizing

• Monitor closely on simmer.markets/dashboard
• Verify trades execute correctly
• Let positions resolve (1-3 days)
• Analyze results

5. Enable Autonomous Trading (Optional)

Only after successful manual testing.

OpenClaw uses metadata-based activation:

• Skills enable via autostart field in SKILL.md
• Verify with OpenClaw docs or platform UI
• Manual editing prevents accidental activation

Steps:

1. Edit SKILL.md line 5 (the metadata line)
2. Change "autostart":false to "autostart":true
3. Save file
4. Restart: openclaw restart
5. Check logs to verify scheduled runs

Manual runs only (alternative):

python weather_trader_enhanced.py --live --smart-sizing

This gives full control over trade timing.

Quick Start

# Check balance
python scripts/status.py

# Execute live trades
python weather_trader_enhanced.py --live --smart-sizing

# Optional: Dry run for testing
python weather_trader_enhanced.py --dry-run

Requirements

• SIMMER_API_KEY from simmer.markets/dashboard → SDK tab
• USDC balance in Simmer wallet

Configuration

Configure via environment variables or config.json:

Variable	Default	Description
SIMMER_WEATHER_ENTRY	0.15	Buy below this price
SIMMER_WEATHER_EXIT	0.45	Sell above this price
SIMMER_WEATHER_MAX_POSITION	5.00	Max USD per trade
SIMMER_WEATHER_MAX_TRADES	5	Max trades per run
SIMMER_WEATHER_LOCATIONS	"ALL"	Cities to target
SIMMER_WEATHER_MIN_QUALITY	0.6	Min market quality

Features

• Dynamic Confidence - Adjusts 60-90% based on forecast lead time
• Market Quality Scoring - Filters low-liquidity markets
• Smart Geocoding - Supports any US city
• Enhanced Parsing - Better temperature bucket detection
• Retry Logic - Recovers from API failures

Commands

# Live trading
python weather_trader_enhanced.py --live --smart-sizing

# Check positions
python weather_trader_enhanced.py --positions

# View config
python weather_trader_enhanced.py --config

# Set config
python weather_trader_enhanced.py --set entry_threshold=0.20

# Optional: Dry run for testing
python weather_trader_enhanced.py --dry-run

How It Works

1. Fetches NOAA forecasts for target locations
2. Discovers weather markets via Simmer API
3. Calculates edge using dynamic confidence model (adjusts 60-90% based on forecast timing)
4. Filters markets by quality score (liquidity, volume, time)
5. Executes trades when price < entry threshold
6. Auto-exits when price > exit threshold

Dynamic confidence adjusts based on lead time:

• Same day: 90%
• 1 day out: 88%
• 2 days out: 85%
• 3 days out: 80%
• 7+ days: 60%

Market quality scoring weights:

• Liquidity (40%)
• Volume (30%)
• Time to resolution (20%)
• Price extremes (10%)

Only trades markets with score ≥ 60%.

Troubleshooting

"SIMMER_API_KEY not set" Set env var or add to .env file

"No weather markets found" Normal if no active markets on Polymarket

"Position size too small" Increase max_position_usd or use --smart-sizing

"Quality score too low" Market filtered for low liquidity/volume

Security Checklist

Before installing, verify:

Code Review

• Inspected weather_trader_enhanced.py source
• Verified 3 network endpoints (NOAA, Nominatim, Simmer)
• Confirmed API key only sent to api.simmer.markets
• No suspicious code or obfuscation

Dependencies

• Verified requirements.txt has only python-dotenv>=1.0.0
• No unused dependencies (web3, telegram, requests removed)
• Using built-in urllib for HTTP
• If installing tradejournal, inspected source first

Configuration

• Confirmed autostart:false in SKILL.md line 5
• Verified activation method with OpenClaw docs/UI
• Know how to disable
• Reviewed position limits ($5 max per trade)

API Key Setup

• Created least-privilege Simmer API key
• Key does NOT have withdrawal permissions
• Key stored in .env file (local, never committed)
• Understand key can read portfolio and place trades

Testing

• Ran python scripts/status.py successfully
• Executed manual trade with small balance using --live
• Monitored trade on simmer.markets/dashboard
• Comfortable with autonomous behavior
• Optional: Tested with --dry-run first

Risk Understanding

• Understand autonomous trading risks
• Know max daily deployment (~$30-60, max $100)
• Comfortable with no per-trade approval
• Know how to stop (disable in OpenClaw or edit SKILL.md)
• Started with small balance ($50-100)

Platform Behavior

• Confirmed OpenClaw respects autostart:false
• Know skill won't run until manually enabled
• Understand cron schedule (every 6 hours when enabled)
• Know how to check OpenClaw logs

If any checkbox is unchecked, DO NOT enable this skill.

Emergency Stop

To stop immediately:

# Option 1: Disable in OpenClaw UI

# Option 2: Edit SKILL.md
# Change line 5: "autostart":false
# Then: openclaw restart

# Option 3: Stop OpenClaw entirely
openclaw stop

# Close positions manually on simmer.markets/dashboard

Support

Source code is visible in weather_trader_enhanced.py All network calls documented in this file No hidden functionality or obfuscation

Use at your own risk. This skill places real trades with real money. Understand prediction markets and trading risks before use.