Back to Documentation
Connect AroAI to your own AI
Integrate AroAI with your preferred AI platform using the Model Context Protocol (MCP)
MCP is the open standard (donated by Anthropic to the Linux Foundation's Agentic AI Foundation in December 2025) that lets AI assistants like Microsoft Copilot, ChatGPT, Claude, and Google Gemini securely call your tools and data. Pick your platform below for the latest setup steps.
MCP Server URL:
Note: The MCP server is not live yet. These instructions are for reference.
https://mcp.aroai.cloud/v1Note: The MCP server is not live yet. These instructions are for reference.
Microsoft Copilot Studio Integration
MCP support in Copilot Studio is now generally available —
connect AroAI in just a few clicks.
- Open Microsoft Copilot Studio and select the agent you want to extend.
- Go to the Tools page for that agent.
- Click Add a tool → New tool → Model Context Protocol.
-
Fill in the server details:
-
Server name:
AroAI -
Server description:
Access AroAI tools for industry-specific intelligence -
Server URL:
https://mcp.aroai.cloud/v1
-
Server name:
- Configure authentication: select OAuth 2.0 and follow the prompts to complete the OAuth flow.
- Pick which tools to expose. All AroAI tools are turned on by default — use the per-tool toggles if you only want a subset.
- Save and authorize the connection. Copilot Studio will auto-sync any new AroAI tools as we publish them — no manual updates needed.
Prerequisite: Generative
Orchestration must be enabled on the agent to use MCP.
Enterprise governance (DLP, VNet, multiple auth
methods) is supported through the underlying connector
infrastructure.
ChatGPT Integration via SSE
Connect ChatGPT Team or Enterprise workspaces directly to
AroAI's cloud-hosted MCP servers using Server-Sent Events
(SSE).
- Open your ChatGPT Workspace Settings and navigate to the "Connected Apps" or "Integrations" page.
- Click 'Add Connection' and select the Model Context Protocol (MCP) option.
-
Configure the connection endpoint:
-
Server Name:
AroAI - Server Type: Select SSE (Server-Sent Events).
-
Server URL:
https://mcp.aroai.cloud/v1/sse
-
Server Name:
- Set the authorization headers: Enter your unique AroAI API key as a Bearer Token (`Authorization: Bearer YOUR_API_KEY`).
- Save and Sync. Once authorized, ChatGPT will read the server's tools schemas. Any team agent can now trigger custom NZ/AU data-retrieval skills during live conversations.
Claude Enterprise Integration
Connect your Claude Enterprise Workspace directly to
AroAI's remote MCP servers. No local installation
required.
- Log in to your Claude Enterprise Admin Console.
- Navigate to Integrations & Tools and select Add Custom MCP Server.
-
Configure the remote server settings:
-
Server Name:
AroAI -
Endpoint URL:
https://mcp.aroai.cloud/v1/sse
-
Server Name:
- Authentication: Select Bearer Token and input your AroAI API key.
- Deploy to Workspace. Once authorized, Claude will instantly fetch the available tools. Your entire organization can now invoke AroAI skills dynamically in any chat without configuring local developer machines.
Google Gemini & Vertex AI Integration
Connect Google Gemini Advanced or Vertex AI Agent Builder
to AroAI's MCP server endpoints.
- Access your Vertex AI or Gemini Developer Console and open your custom agent project.
- Navigate to 'Extensions' and click 'Add Custom Extension'.
- Select Model Context Protocol (MCP) as the extension type.
-
Provide the connection parameters:
-
Extension Name:
AroAI_Skills -
Endpoint URL:
https://mcp.aroai.cloud/v1/sse
-
Extension Name:
- Configure Authentication: Select Bearer Token and insert your AroAI API key.
- Publish the Extension. Gemini's agentic orchestrator will dynamically call our pre-processing and retrieval tools to verify registries or fetch regional regulations during customer queries.
Custom Agent & SDK Configurations
Connect custom Python or TypeScript AI agents to AroAI
using standard MCP SDK libraries.
Python SDK Integration
import asyncio
from mcp.client.sse import sse_client
from mcp.client.session import ClientSession
async def main():
# Connect via Server-Sent Events (SSE) to AroAI's hosted endpoint
headers = {"Authorization": "Bearer YOUR_API_KEY"}
url = "https://mcp.aroai.cloud/v1/sse"
async with sse_client(url, headers=headers) as streams:
async with ClientSession(streams[0], streams[1]) as session:
await session.initialize()
# Execute a standard retrieval skill
result = await session.call_tool(
"nz_company_research",
{"query": "AroAI Limited"}
)
print(result.content)
asyncio.run(main())
TypeScript SDK Integration
import { Client } from "@modelcontextprotocol/sdk/client/index.js";
import { SSEClientTransport } from "@modelcontextprotocol/sdk/client/sse.js";
const client = new Client({
name: "custom-enterprise-agent",
version: "1.0.0"
}, {
capabilities: {}
});
// Use SSE transport for remote HTTP connections
const transport = new SSEClientTransport(
new URL("https://mcp.aroai.cloud/v1/sse"),
{ headers: { Authorization: "Bearer YOUR_API_KEY" } }
);
await client.connect(transport);
// Dynamically list and trigger regional skills
const tools = await client.listTools();
console.log("Exposed Tools:", tools);
