{"service": "ArbiIntel", "version": "5.0.0", "description": "Cryptocurrency arbitrage intelligence engine", "mcp_endpoint": "POST /mcp", "tools": [{"name": "funding_rates", "description": "Funding rate arbitrage scanner across 7 CEXs (OKX, Binance, Bybit, HTX, Gate.io, MEXC, Crypto.com). Finds the best carry trade opportunities by comparing funding rates across exchanges.\n\nUSE CASE: When you need to identify which exchange offers the highest funding rate for a specific token, or when funding rate discrepancies exceed 0.01% between exchanges.\n\nOUTPUT: List of opportunities sorted by spread_percent (descending). Each includes: token, exchange_A_rate, exchange_B_rate, spread, annualized_roi_pct, confidence, weighted_score (0-100).\n\nEXAMPLE: 'BTC on Binance (0.02%) vs OKX (-0.01%) = 0.03% spread → 26.28% annualized ROI'", "inputSchema": {"type": "object", "properties": {}, "required": []}}, {"name": "dex_cex_arb", "description": "Cross-chain CEX-DEX arbitrage scanner. Compares prices of wrapped tokens (WBTC, WETH, etc.) across 8+ blockchain networks (ETH, ARB, OPT, POL, BASE, BSC, SOL, SUI) between DEX and CEX.\n\nUSE CASE: When you need to find price discrepancies between decentralized exchanges (Uniswap, SushiSwap, PancakeSwap) and centralized exchanges (OKX, Binance) for the same token.\n\nOUTPUT: Opportunities sorted by spread_percent (descending). Each includes: token, chain, dex_price, cex_price, spread, liquidity_usd, volume_24h, weighted_score.\n\nEXAMPLE: 'WBTC on Arbitrum ($62,341) vs OKX ($62,418) = 0.12% spread'", "inputSchema": {"type": "object", "properties": {"chain": {"type": "string", "description": "Blockchain networks supported: ethereum, arbitrum, optimism, polygon, base, binance, solana, sui, avalanche, tron, near, aptos, ton, zkysnc, blast, linea, scroll, starknet, fantom, xrp, ada. Default: ethereum", "default": "ethereum"}, "limit": {"type": "integer", "description": "Maximum results to return (1-50). Default: 10", "default": 10}}, "required": []}}, {"name": "triangular_arb", "description": "Triangular arbitrage scanner. Identifies price discrepancies between 3 consecutive trades (e.g., USDT→BTC→ETH→USDT).\n\nUSE CASE: When you want to exploit pricing inefficiencies across multiple trading pairs on the same exchange. Works best during high volatility periods.\n\nOUTPUT: Opportunities sorted by spread_percent. Each includes: path, direction, spread, confidence, weighted_score.\n\nEXAMPLE: 'USDT→BTC→ETH→USDT = 0.35% spread → $0.35 profit per $100 traded'", "inputSchema": {"type": "object", "properties": {}, "required": []}}, {"name": "hot_tokens", "description": "Trending token scanner. Identifies tokens with unusual 1-hour trading volume spikes (>$10k volume, >$50k liquidity).\n\nUSE CASE: When you need to detect emerging market trends, new trading opportunities, or potential pump-and-dump schemes before they happen.\n\nOUTPUT: Tokens sorted by volume_1h (descending). Each includes: token_address, price_usd, volume_1h, liquidity_usd, price_change_1h, weighted_score, risk_level.\n\nEXAMPLE: 'PEPE on Ethereum: $2.3M volume (15.8% price change in 1h), score: 90/100'", "inputSchema": {"type": "object", "properties": {"chain": {"type": "string", "description": "Blockchain networks supported: ethereum, arbitrum, optimism, polygon, base, binance, solana, sui, avalanche, tron, near, aptos, ton, zkysnc, blast, linea, scroll, starknet, fantom, xrp, ada. Default: ethereum", "default": "ethereum"}, "limit": {"type": "integer", "description": "Maximum tokens to return (1-50). Default: 10", "default": 10}}, "required": []}}, {"name": "cross_chain_bridge", "description": "Cross-chain bridge arbitrage scanner. Compares prices of wrapped tokens (WBTC, wstETH, stETH, renBTC, tBTC) across 5+ chains.\n\nUSE CASE: When you need to identify bridge inefficiencies between Layer 1 and Layer 2 networks. Especially useful for large capital deployments.\n\nOUTPUT: Opportunities sorted by spread_percent. Each includes: token, chain_a, chain_b, price_a, price_b, spread, liquidity, bridge_risk, weighted_score.\n\nEXAMPLE: 'WBTC on Ethereum ($62,341) vs Arbitrum ($62,418) = 0.12% spread'", "inputSchema": {"type": "object", "properties": {}, "required": []}}, {"name": "liquidation_alerts", "description": "Extreme funding rate alert system. Detects funding rates >±0.5%/8h that may trigger liquidation cascades.\n\nUSE CASE: When you need early warnings of potential market crashes, long/short squeezes, or mass liquidation events.\n\nOUTPUT: Alerts sorted by risk_level (critical → high → medium). Each includes: token, avg_funding_rate, max_funding_rate, exchanges_covered, estimated_open_interest, signal, weighted_score.\n\nEXAMPLE: 'SOL on Crypto.com: 0.05% funding rate (last 3 cycles) → LONG SQUEEZE RISK, score: 68/100'", "inputSchema": {"type": "object", "properties": {}, "required": []}}, {"name": "new_token_scanner", "description": "新币扫描器 — 扫描 OKX 永续合约市场上最近上线的交易对和 24 小时内成交量异动的代币。\n\n适用场景：发现 OKX 刚上线的新币种，或捕捉 24 小时内价格变化超 100% 的代币。适合抢先布局趋势币。\n\n输出：两份清单 — (a) 最近 7 天内新上线的币种，(b) 24 小时内涨幅超 100% 的代币。含价格、成交量、风险等级。\n\n示例：'ATOMPERP 上线仅 2 小时, 成交量 $2.3M, 风险：极高'", "inputSchema": {"type": "object", "properties": {"limit": {"type": "integer", "description": "每个分类最多返回条数（默认20）。", "default": 20}}, "required": []}}, {"name": "whale_alert", "description": "巨鲸异动监控 — 追踪以太坊上知名巨鲸钱包的大额转账（>$1M USD），覆盖 USDT/USDC/WBTC/DAI。\n\n适用场景：提前感知大资金动向——巨鲸增持往往是价格走势的先行信号。\n\n输出：按 USD 金额降序排列的巨鲸转账列表，含方向(转入/转出)、代币、数量、预估 USD 价值、风险信号。\n\n示例：'Coinbase CEO 钱包收到 5M USDT(转入) — 看涨信号'", "inputSchema": {"type": "object", "properties": {"wallet_address": {"type": "string", "description": "指定监控的钱包地址（hex 格式）。留空则监控所有已追踪的巨鲸。"}, "limit": {"type": "integer", "description": "Max alerts per whale (default 10).", "default": 10}}, "required": []}}, {"name": "gas_price_tracker", "description": "Gas 费用追踪 — 追踪以太坊链上 Gas 价格，帮你找到 DEX 交易的最优时机。计算扣除 Gas 后的实际套利利润。\n\n适用场景：决定「现在就执行这笔 DEX 套利，Gas 够便宜吗？」它会告诉你答案。\n\n输出：当前 Gas 价格、等级(低/中/高/极高)、ETH 价格、质押 APR、交易盈利分析和可执行建议。\n\n示例：'Gas: 12 Gwei(低) — 黄金窗口！Swap 成本仅 $0.63。$100 套利净利: $99.37'", "inputSchema": {"type": "object", "properties": {"expected_profit": {"type": "number", "description": "预期套利利润(USD)，用于计算交易盈利能力（默认100）。", "default": 100.0}}, "required": []}}, {"name": "stablecoin_depeg_monitor", "description": "稳定币脱锚监控 — 跨多条链实时监控 USDT/USDC/DAI/FDUSD/PYUSD 的锚定稳定性。一旦跌破 $0.98 立刻告警。\n\n适用场景：在做大额 DeFi 仓位前快速评估稳定币风险。脱锚事件往往是更大市场危机的先行信号。\n\n输出：每个稳定币的当前价格、锚定比率、24h 变化，低于 $0.98 触发 ⚠️ 警告。含整体风险等级和行动建议。\n\n示例：'USDT: $0.9998(健康), USDC: $0.9995(健康), DAI: $0.9790 ⚠️ 脱锚警告'", "inputSchema": {"type": "object", "properties": {"chain": {"type": "string", "description": "指定监控的链（可选）。留空则检查所有链。"}}, "required": []}}, {"name": "yield_spread", "description": "Compare DeFi lending yields across Aave, Compound, Venus, and other protocols on 15+ chains. Detects cross-protocol basis trade opportunities on stablecoin deposits.\n\nUSE CASE: When you need to identify which DeFi lending protocol offers the best yield for stablecoins (USDT, USDC, DAI) across multiple chains. Helps find basis trades where lending yields differ significantly between protocols.\n\nOUTPUT: Yield comparison across protocols and chains. Each includes: protocol, chain, supply_apy, borrow_apy, total_supplied_usd, utilization_rate, risk_adjusted_apy.\n\nEXAMPLE: 'USDC on Aave v3 (Ethereum) 4.2% APY vs Compound v3 (Polygon) 6.8% APY = 2.6% basis spread'", "inputSchema": {"type": "object", "properties": {"chains": {"type": "array", "items": {"type": "string"}, "description": "Filter by chains (e.g., ['ethereum', 'polygon', 'avalanche']). Empty for all chains."}, "stablecoins": {"type": "array", "items": {"type": "string"}, "description": "Filter by stablecoins (e.g., ['USDC', 'USDT', 'DAI']). Empty for all."}, "limit": {"type": "integer", "description": "Maximum results (default 20).", "default": 20}}, "required": []}}, {"name": "mev_flash_swap", "description": "Detect potential MEV (Maximal Extractable Value) opportunities across major DEX aggregators: flash swap arbitrages, sandwich attack vectors, and liquidation rebates.\n\nUSE CASE: When you need to identify MEV opportunities in the mempool or across DEX routes. Covers 1inch, 0x, CowSwap, and Uniswap Router scans for profitable atomic transactions.\n\nOUTPUT: Detected MEV opportunities. Each includes: type (flash_swap/sandwich/liquidation), estimated_profit_usd, gas_cost_usd, net_profit_usd, risk_level, execution_complexity.\n\nEXAMPLE: 'Flash swap: WBTC/ETH on Uniswap → Sushiswap arb, est. profit $234, risk: medium'", "inputSchema": {"type": "object", "properties": {"min_profit": {"type": "number", "description": "Minimum estimated profit in USD to report (default 10).", "default": 10}, "opportunity_types": {"type": "array", "items": {"type": "string"}, "description": "Types to detect: flash_swap, sandwich, liquidation. Empty for all."}, "limit": {"type": "integer", "description": "Maximum results (default 15).", "default": 15}}, "required": []}}, {"name": "raw_funding_rates", "description": "RAW, un-scored funding rates across OKX/Binance/Bybit (top_n by absolute annualized rate). For agents that want to compute their own strategies instead of using our scored output.\n\nUSE CASE: When you need the underlying per-exchange funding rate feed to build custom carry or basis models.\n\nOUTPUT: data[] with {exchange, token, rate_pct, annualized, source}. No scoring applied.", "inputSchema": {"type": "object", "properties": {"top_n": {"type": "integer", "description": "Number of top rates to return by abs annualized (default 20).", "default": 20}}, "required": []}}, {"name": "raw_dex_prices", "description": "RAW DEX pair prices for a chain (Ethereum/Arbitrum/Base/Solana/BSC), un-scored. For composing your own CEX-DEX or cross-chain arbitrage.\n\nUSE CASE: When you need raw on-chain DEX liquidity + price data to cross-check against CEX or build bespoke strategies.\n\nOUTPUT: data[] with {chain, dex, base, quote, price_usd, liquidity_usd, volume_24h, source}.", "inputSchema": {"type": "object", "properties": {"chain": {"type": "string", "description": "Chain: ethereum, arbitrum, base, solana, bsc. Default: ethereum", "default": "ethereum"}, "limit": {"type": "integer", "description": "Max pairs to return (default 20).", "default": 20}}, "required": []}}, {"name": "day_pass", "description": "Returns ArbiIntel day-pass status and pricing. A day-pass unlocks many calls for one x402 payment. Currently per-call pricing is active at $0.03; this tool reports when day-pass is enabled and its terms.\n\nOUTPUT: {enabled, price_usd, calls_included, window_hours, note}.", "inputSchema": {"type": "object", "properties": {}, "required": []}}, {"name": "best_arbitrage", "description": "AGENT DECISION ENGINE (top-level). Given a principal and risk appetite, it consumes the underlying arbitrage scanners (funding carry, CEX-DEX, cross-chain bridge) and returns a single DECISION an agent can execute directly — NOT raw data.\n\nINPUT: principal (USD, default 1000), risk_appetite (conservative|balanced|aggressive, default balanced), chain (default ethereum), limit.\n\nOUTPUT: decision (EXECUTE|WAIT|SKIP), opportunity_score(0-100), decision_confidence(%) with confidence_factors (market_agreement / liquidity / risk_stability / historical_pattern), expected_profit_usd, risk_level, ttl_minutes, reason, decision_explanation {why[], risk_notes[]}, top_opportunity, alternatives[], market_context.\n\nUSE CASE: a trading agent asks 'what should I do right now?' and calls this ONCE instead of composing 15 tools. The engine also decides NOT to act when the edge is thinner than gas (see SKIP cases).\n\nEXAMPLE: principal=1000, balanced → decision=EXECUTE, SOL funding carry, spread 612% annualized, decision_confidence 88% (led by market_agreement), TTL 336m, est. $5.59 net, decision_explanation={why:[spread expanded, confidence high, gas cleared], risk:[reversion within TTL]}.", "inputSchema": {"type": "object", "properties": {"principal": {"type": "number", "description": "Capital to deploy in USD (default 1000).", "default": 1000}, "risk_appetite": {"type": "string", "description": "conservative | balanced | aggressive (default balanced).", "default": "balanced"}, "chain": {"type": "string", "description": "Chain for CEX-DEX scan (default ethereum).", "default": "ethereum"}, "limit": {"type": "integer", "description": "Max results per scanner (default 10).", "default": 10}}, "required": []}}, {"name": "decision_accuracy", "description": "Returns the hub's REAL, backtest-verified decision accuracy (no hardcoded numbers). Use to audit how often EXECUTE recommendations held their edge. Free to call.", "inputSchema": {"type": "object", "properties": {}}}]}