About Arbilink
ArbiLink is an MCP server for AI agents working with Arbitrum, offering on-chain data queries and DeFi transaction preparation. It includes GMX V2 position simulations, Uniswap V3 swap data, and an on-chain AgentRegistry.
Key features
- 01Query balances, portfolios, and transactions
- 02Read blocks, gas prices, and DeFi rates
- 03Simulate GMX V2 perpetual positions
- 04Prepare Uniswap V3 swap data
- 05Check GMX position health
Use cases
- 01Inspect assets at an Arbitrum address
- 02Assess risk in a simulated GMX leveraged position
- 03Prepare a Uniswap token swap
- 04Check Aave V3 lending and borrowing rates
MCP support
It exposes 11 tools to AI agents through MCP JSON-RPC 2.0 and also provides a REST API. Transaction tools simulate or prepare transaction data; the README does not describe server-side signing or execution.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
Tools
Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.
Details
From the project READMEArbiLink — Arbitrum MCP Server for AI Agents
ArbiLink is a native Model Context Protocol (MCP) server that connects AI agents directly to the Arbitrum blockchain. Claude, Cursor, Windsurf, and any MCP-compatible agent can read live on-chain data and prepare DeFi transactions without writing a single line of blockchain code.
"Your AI agent is blockchain-blind. ArbiLink gives it eyes."
Live Endpoints
| Interface | URL | Protocol |
|---|---|---|
| Native MCP | https://arbi-link-challenge.replit.app/mcp | JSON-RPC 2.0 (MCP 2024-11-05) |
| REST API | https://arbi-link-challenge.replit.app/api/mcp | HTTP |
| Dashboard | https://arbi-link-challenge.replit.app | Web |
Quick Start — Claude Desktop
Add to your claude_desktop_config.json:
{
"mcpServers": {
"arbilink": {
"command": "npx",
"args": [
"mcp-remote",
"https://arbi-link-challenge.replit.app/mcp"
]
}
}
}
Then ask Claude:
- "What's in vitalik.eth's Arbitrum wallet?"
- "Simulate a 10x ETH long with $500 collateral on GMX"
- "Prepare a swap of 1 ETH to USDC on Uniswap V3"
- "What's the current gas price on Arbitrum?"
Quick Start — Cursor / Windsurf
Add to your MCP settings:
{
"mcpServers": {
"arbilink": {
"url": "https://arbi-link-challenge.replit.app/mcp"
}
}
}
11 Live Tools
Read Tools (8) — Live on-chain data
| Tool | Description |
|---|---|
get_eth_balance | ETH balance for any Arbitrum address |
get_token_balance | ERC-20 balance (ARB, USDC, USDT, GMX, LINK, WBTC, and more) |
get_wallet_portfolio | Full portfolio with USD values and 24h change |
get_transaction | Decode any transaction by hash |
get_gas_price | Live gas price on Arbitrum One |
get_block | Block data by number or latest |
get_contract_abi | Fetch and decode ABI from Arbiscan |
get_defi_rates | Live Aave V3 lending and borrowing rates |
Execution Tools (3) — Simulate and prepare before signing
| Tool | Description |
|---|---|
simulate_gmx_open | Full GMX V2 perpetual simulation — entry price, liquidation distance, PnL table at ±5/10/20%, hourly borrow fee |
prepare_uniswap_swap | Uniswap V3 swap quote + complete exactInputSingle calldata ready to submit to SwapRouter02 |
get_gmx_position_health | Live open GMX V2 positions for any wallet — size, unrealized PnL, health status (HEALTHY / AT RISK / CRITICAL) |
Example Tool Calls
Simulate a GMX position
{
"jsonrpc": "2.0",
"id": 1,
"method": "tools/call",
"params": {
"name": "simulate_gmx_open",
"arguments": {
"indexToken": "ETH",
"collateralUsd": 500,
"leverage": 10,
"direction": "long"
}
}
}
Response includes: live ETH price, position size, liquidation price (with % distance), PnL at 6 price scenarios, hourly borrow fee, and exact next steps to open the position on GMX V2.
Prepare a Uniswap swap
{
"jsonrpc": "2.0",
"id": 2,
"method": "tools/call",
"params": {
"name": "prepare_uniswap_swap",
"arguments": {
"tokenIn": "ETH",
"tokenOut": "USDC",
"amountIn": 1.5
}
}
}
Response includes: live quote, estimated output, slippage-protected minimum, pool fee tier, price impact, gas estimate, and complete exactInputSingle calldata for SwapRouter02.
Check wallet portfolio
{
"jsonrpc": "2.0",
"id": 3,
"method": "tools/call",
"params": {
"name": "get_wallet_portfolio",
"arguments": {
"address": "0xd8dA6BF26964aF9D7eEd9e03E53415D37aA96045"
}
}
}
REST API
All tools are also available via HTTP POST:
curl -X POST https://arbi-link-challenge.replit.app/api/mcp \
-H "Content-Type: application/json" \
-d '{
"tool": "get_eth_balance",
"arguments": { "address": "0xd8dA6BF26964aF9D7eEd9e03E53415D37aA96045" }
}'
On-Chain Agent Registry
ArbiLink ships with an AgentRegistry smart contract deployed on Arbitrum Sepolia — a verifiable on-chain identity primitive for AI agents.
- Contract:
0x8fb1a3fe1e380b5578918888140a16fa699c85b0(Arbitrum Sepolia) - Agents register their endpoint, capabilities, and metadata on-chain
- Any client can verify an agent's identity and capabilities trustlessly
- Inspectable via the Agent page
Supported Tokens
ETH, WETH, ARB, USDC, USDT, WBTC, GMX, LINK, DAI, PENDLE, RDNT
Supported DeFi Protocols
- GMX V2 — perpetual simulation, position health
- Uniswap V3 — swap preparation with calldata
- Aave V3 — live lending/borrowing rates
Architecture
AI Agent (Claude / Cursor / Windsurf)
│
▼ MCP JSON-RPC 2.0
┌─────────────────────┐
│ ArbiLink Server │
│ (Node.js/Express) │
├─────────────────────┤
│ 11 MCP Tools │
│ REST API bridge │
│ AgentRegistry RPC │
└─────────────────────┘
│
▼
┌─────────────────────┐
│ Arbitrum One │
│ (live RPC) │
│ CoinGecko prices │
│ GMX V2 API │
│ Arbiscan ABI │
└─────────────────────┘
Roadmap
- v1.0 (Live) — 8 read tools + 3 execution tools, AgentRegistry contract, dual MCP/REST interfaces
- v1.1 (Building) — API key auth + rate limiting, Aave V3 positions, Uniswap LP positions
- v2.0 (Planned) — Multi-agent coordination, webhook subscriptions for on-chain events
- v3.0 (Future) — Transaction signing via connected wallet, multi-chain expansion
License
MIT
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.