For AI Agents

Quick Start

Get VAEA Flash running in Claude Desktop or Cursor in under 2 minutes.

Claude Desktop

Add to your claude_desktop_config.json:

json

{
  "mcpServers": {
    "vaea-flash": {
      "command": "npx",
      "args": ["-y", "@vaea/mcp-server"],
      "env": {
        "VAEA_API_URL": "https://api.vaea.fi"
      }
    }
  }
}

Restart Claude Desktop. You'll see the VAEA tools in the 🔨 menu. Try: "What SOL liquidity is available for a flash loan?"

Cursor

Add to .cursor/mcp.json in your project:

json

{
  "mcpServers": {
    "vaea-flash": {
      "command": "npx",
      "args": ["-y", "@vaea/mcp-server"]
    }
  }
}

First Tool Call

Once connected, ask your AI agent:

💬 "What flash loan liquidity is available on Solana right now?"

💬 "Find the cheapest route to borrow 1000 SOL"

💬 "Is a 1000 SOL flash loan profitable if I expect 0.5 SOL revenue?"

💬 "Build me a flash loan transaction for 500 USDC"

💡 Tip

The agent will automatically choose the right MCP tool based on your question. You don't need to know the tool names — just describe what you want in natural language.

Rust Server

For lower memory and faster startup, use the native Rust binary:

json

{
  "mcpServers": {
    "vaea-flash": {
      "command": "/path/to/vaea-mcp",
      "env": {
        "VAEA_API_URL": "https://api.vaea.fi"
      }
    }
  }
}

Build with cargo build --release in the mcp/rust/ directory. The binary is ~4MB and starts in milliseconds.