WhaleWatch CLI — Agent Skill

Track crypto whale movements. Close the loop from on-chain signal → agent reasoning → prediction market bet.

Install

uv pip install whalecli

Quick Start

# Initialize config (creates ~/.config/whalecli/config.toml)
whalecli config init

# Set API key (free tier: 5 req/sec)
whalecli config set api.etherscan_api_key YOUR_KEY

# Add a whale wallet
whalecli wallet add 0xd8dA6BF26964aF9D7eEd9e03E53415D37aA96045 --label "vitalik.eth" --chain ETH

# Scan all wallets (last 24h)
whalecli scan --format json

# Stream real-time alerts (JSONL, one event per line)
whalecli stream --chain ETH --interval 60

When to Use

Trigger phrases:

• "What are the whales doing?"
• "Check whale activity on ETH/BTC"
• "Are whales accumulating or distributing?"
• "Any large wallet movements?"
• "What's the on-chain signal?"
• "Run a whale scan"

Automatic triggers (no user prompt needed):

• Every 60 minutes during market-active hours (via heartbeat)
• Before placing a Simmer/Polymarket bet (pre-bet whale sanity check)
• When fear/greed index crosses a threshold (corroboration)

CLI Commands

whalecli scan — One-shot whale analysis

whalecli scan --chain ETH --hours 4 --threshold 70 --format json

Output (JSON):

{
  "scan_id": "scan_20260222_103015_a1b2",
  "chain": "ETH",
  "window_hours": 4,
  "wallets": [
    {
      "address": "0xd8dA...",
      "label": "vitalik.eth",
      "score": 82,
      "direction": "accumulating",
      "score_breakdown": {
        "net_flow": 35,
        "velocity": 20,
        "correlation": 15,
        "exchange_flow": 12
      },
      "net_flow_usd": 15000000,
      "tx_count": 12
    }
  ],
  "summary": {
    "total_wallets": 5,
    "accumulating": 3,
    "distributing": 1,
    "neutral": 1,
    "avg_score": 65
  },
  "alerts_triggered": 2
}

whalecli stream — Real-time JSONL streaming

whalecli stream --chain ETH --interval 60 --threshold 70

Events (one JSON per line):

• stream_start — stream initialized
• whale_alert — score exceeded threshold
• whale_activity — activity detected below threshold
• heartbeat — periodic health check
• stream_end — clean shutdown

whalecli wallet — Manage tracked wallets

whalecli wallet add 0x... --label "whale1" --chain ETH
whalecli wallet list --format json
whalecli wallet remove 0x...
whalecli wallet import wallets.csv

whalecli alert — Configure alert rules

whalecli alert set --score 75 --webhook https://example.com/hook
whalecli alert set --threshold 1000000 --window 1h
whalecli alert list --format json

whalecli report — Historical analysis

whalecli report --wallet 0x... --days 30 --format json

whalecli config — Configuration management

whalecli config init
whalecli config set api.etherscan_api_key YOUR_KEY
whalecli config show

Exit Codes

Scoring Algorithm

4-dimension whale scoring (0–100 points):

• Net Flow (0–40): USD net flow with log10 scaling, wallet age weighting
• Velocity (0–25): Current activity vs 30-day baseline (log2 ratio)
• Correlation (0–20): Peer wallet direction agreement (min 2 peers)
• Exchange Flow (0–15): CEX address registry lookup + direction match

Interpretation:

• 80–100: Strong whale signal (high confidence)
• 60–79: Moderate activity (worth monitoring)
• 40–59: Low activity (noise)
• 0–39: Minimal (ignore)

Agent Integration Pattern

import subprocess, json

def whale_scan(chain="ETH", hours=4, threshold=70):
    """Run whale scan and return parsed results."""
    result = subprocess.run(
        ["whalecli", "scan", "--chain", chain,
         "--hours", str(hours), "--threshold", str(threshold),
         "--format", "json"],
        capture_output=True, text=True, timeout=30
    )
    if result.returncode == 2:
        raise RuntimeError(f"API error: {result.stderr}")
    if not result.stdout.strip():
        return {"wallets": [], "alerts_triggered": 0}
    return json.loads(result.stdout)

# Example: pre-bet whale check
scan = whale_scan(chain="ETH", hours=4)
if scan["summary"]["accumulating"] > scan["summary"]["distributing"]:
    print("Whales accumulating — bullish signal")

FearHarvester Integration

The closed loop: Fear & Greed → Whale Signal → Simmer Bet

# 1. Get F&G value
fg_value = get_fear_greed_index()  # e.g., 8 (Extreme Fear)

# 2. Check whale confirmation
scan = whale_scan(chain="ETH", hours=4)
whales_accumulating = scan["summary"]["accumulating"] > scan["summary"]["distributing"]

# 3. If fear + whales accumulating → strong contrarian signal
if fg_value <= 20 and whales_accumulating:
    # Place bet on recovery market
    place_simmer_bet(market="btc_recovery", side="yes", amount=15)

Supported Chains

• ETH — Etherscan API (free tier: 5 req/sec)
• BTC — Mempool.space (primary) + Blockchain.info (fallback)
• HL — Hyperliquid perpetual fills and positions

Links

• PyPI: https://pypi.org/project/whalecli/
• GitHub: https://github.com/clawinfra/whalecli
• Issues: https://github.com/clawinfra/whalecli/issues