ProfitSpot MCP
以下摘录自该仓库的 README(片段) · 在 GitHub 读完整 README →
Cross-chain DeFi intelligence for AI agents.
Risk-scored yields, Monte Carlo simulations, whale tracking
across 86 chains and 6,500+ pools.
Quick Install •
Examples •
Tool Reference •
Pricing •
Self-Hosting
---
What is this?
ProfitSpot MCP gives any AI agent — Claude, GPT, Gemini, or your custom agent — instant access to the same DeFi intelligence that powers profitspot.live. Think of it as the **Bloomberg Terminal for AI agents**.
Existing DeFi MCP servers are either basic DeFiLlama wrappers with 1 tool, or locked to a single chain. ProfitSpot MCP covers **86 chains**, **6,500+ pools**, and adds an **intelligence layer** — risk grades, Monte Carlo profit simulation, whale detection — that no competitor has.
**No API key required to get started.** Free tier gives you yield discovery, market overview, and IL calculations immediately.
---
🚀 Quick Install
Claude Desktop
Add to your claude_desktop_config.json:
macOS ~/Library/Application Support/Claude/claude_desktop_config.json
Windows %APPDATA%\Claude\claude_desktop_config.json
> **Free tier?** Just remove the "env" block entirely. No key needed.
Cursor
Add to .cursor/mcp.json in your project root:
VS Code (Copilot)
Add to your .vscode/mcp.json:
pip (manual install)
---
💡 Example Tool Calls
Real responses from live DeFiLlama data. Every response includes confidence (0–1) and data_freshness (ISO timestamp).
1. Discover Yields — Ethereum, grade B or better
Full JSON Response (click to expand)
**What you get:** 146 Ethereum pools matched from 16,642 scanned. Each pool comes risk-scored (A–F), with 30-day mean APY and APY prediction direction. Sorted by APY descending.
---
2. Risk Score — Deep breakdown for a Uniswap V3 pool
Full JSON Response (click to expand)
**What you get:** A composite **B grade (71/100)** with 6-factor breakdown. This pool scores perfectly on protocol reputation (Uniswap is Tier 1 Blue Chip, 48+ months) and security, but gets dinged on APY volatility (CV=1.04) and non-stablecoin pair risk. An AI agent can read this and explain *exactly why* a pool got its grade.
---
3. Impermanent Loss — ETH +50%, USDC −10%
Full JSON Response (click to expand)
**What you get:** Exact IL of **3.18% ($381)** on a $10K position. Your LP is worth $11,619 vs $12,000 if you just held. But here's the key insight — the net_with_fees table shows that even a modest 5% fee APY makes you net **+$2,119 (+21.2% ROI)**. The tool gives your AI agent the full picture to make an informed recommendation.
---
📖 Tool Reference
All 7 tools available through the MCP protocol:
discover_yields 🟢 Free
> Find top yield opportunities across 86 chains. Filter, risk-score, and rank.
| Parameter | Type | Default | Description |
|-----------|------|---------|-------------|
| chain | string \| null | null | Filter by chain name (e.g., "Ethereum", "Arbitrum") |
| min_tvl | float | 1000000 | Minimum TVL in USD |
| min_apy | float | 0 | Minimum APY percentage |
| max_risk | string | "F" | Maximum risk grade: A, B, C, D, or F *(Pro only)* |
| limit | int | 20 | Max results to return (Free: 10, Pro: 50) |
---
analyze_pool 🔒 Pro
> Deep-dive analysis of a single pool: APY trends, risk breakdown, IL scenarios, projected returns.
| Parameter | Type | Default | Description |
|-----------|------|---------|-------------|
| pool_id | string | *required* | Pool UUID from discover_yields results |
**Returns:** Historical APY (7d/30d/90d averages), trend direction, volatility stdev, full risk grade breakdown, IL estimate for 3 divergence scenarios, and projected returns for $1K/$10K/$100K at 30/90/365 days.
---
calculate_impermanent_loss 🟢 Always Free
> Pure math — exact IL for any 50/50 LP pair. No API call needed.
| Parameter | Type | Default | Description |