Setup Guide¶

Prerequisites¶

• An active ChainAware API key — get one at chainaware.ai/support or from your profile page
• An MCP-compatible client (Claude Code, Claude Desktop, Cursor, or any MCP SDK)

---

Claude Code¶

The fastest path. One command registers the server globally across all your Claude Code sessions.

claude mcp add --transport sse chainaware-behavioral-prediction \
  https://prediction.mcp.chainaware.ai/sse \
  --header "X-API-Key: YOUR_API_KEY"

Verify the server is registered:

claude mcp list

You should see chainaware-behavioral-prediction in the list. The six tools are now available in any Claude Code session.

Installing the Subagents (Recommended)¶

The repository includes 31 pre-built subagents. Copy them into your project's .claude/agents/ directory:

git clone https://github.com/ChainAware/behavioral-prediction-mcp.git
cp behavioral-prediction-mcp/.claude/agents/*.md .claude/agents/

Invoke any agent by name within a Claude Code session:

@chainaware-fraud-detector check 0xd8dA6BF26964aF9D7eEd9e03E53415D37aA96045 on ETH

---

Claude Desktop¶

Add the server to your Claude Desktop MCP configuration file.

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "chainaware-behavioral-prediction": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "https://prediction.mcp.chainaware.ai/sse",
        "--header",
        "X-API-Key: YOUR_API_KEY"
      ]
    }
  }
}

Restart Claude Desktop after saving the file. The ChainAware tools will appear in the tools panel.

---

Cursor¶

Add the server to your Cursor MCP settings at Cursor → Settings → MCP.

{
  "mcpServers": {
    "chainaware-behavioral-prediction": {
      "url": "https://prediction.mcp.chainaware.ai/sse",
      "headers": {
        "X-API-Key": "YOUR_API_KEY"
      }
    }
  }
}

---

Node.js / Custom MCP Client¶

Install the MCP SDK and connect to the server programmatically:

npm install @modelcontextprotocol/sdk

import { Client } from "@modelcontextprotocol/sdk/client/index.js";
import { SSEClientTransport } from "@modelcontextprotocol/sdk/client/sse.js";

const transport = new SSEClientTransport(
  new URL("https://prediction.mcp.chainaware.ai/sse"),
  {
    requestInit: {
      headers: { "X-API-Key": process.env.CHAINAWARE_API_KEY }
    }
  }
);

const client = new Client({ name: "my-app", version: "1.0.0" }, {});
await client.connect(transport);

// Call a tool
const result = await client.callTool({
  name: "predictive_fraud",
  arguments: {
    apiKey: process.env.CHAINAWARE_API_KEY,
    network: "ETH",
    walletAddress: "0xd8dA6BF26964aF9D7eEd9e03E53415D37aA96045"
  }
});

console.log(result.content[0].text);

---

Python / Custom MCP Client¶

pip install mcp

import asyncio
import os
from mcp import ClientSession
from mcp.client.sse import sse_client

async def main():
    async with sse_client(
        url="https://prediction.mcp.chainaware.ai/sse",
        headers={"X-API-Key": os.environ["CHAINAWARE_API_KEY"]}
    ) as (read, write):
        async with ClientSession(read, write) as session:
            await session.initialize()

            result = await session.call_tool("predictive_fraud", {
                "apiKey": os.environ["CHAINAWARE_API_KEY"],
                "network": "ETH",
                "walletAddress": "0xd8dA6BF26964aF9D7eEd9e03E53415D37aA96045"
            })
            print(result.content[0].text)

asyncio.run(main())

---

Environment Variables¶

Always store your API key as an environment variable. Never hardcode it in source files or commit it to version control.

# .env (add to .gitignore)
CHAINAWARE_API_KEY=your-api-key-here

# Shell
export CHAINAWARE_API_KEY="your-api-key-here"

---

Verifying the Connection¶

Once connected, test the setup by calling predictive_fraud on a known address:

Check fraud risk for vitalik.eth on Ethereum

A successful response returns a probabilityFraud score and status field. If you receive a 403 error, check that your API key is correctly set. If you receive a connection error, verify the MCP server URL is https://prediction.mcp.chainaware.ai/sse.

---

Troubleshooting¶

---

See also: Prediction MCP Overview | predictive_fraud | GitHub Repository