Other AI Agents & Agentic Frameworks

The eConnect MCP Server uses the standard Model Context Protocol (MCP) with Streamable HTTP transport — the same protocol supported by a growing ecosystem of AI tools, agentic frameworks, and orchestration platforms. If your tool supports MCP over HTTP, it can connect to eConnect.

MCP Endpoint:

https://mcp.econnect.ai/mcp

Transport type: Streamable HTTP (JSON-RPC over HTTP POST)

---

Generic MCP Client Setup

For any MCP-compatible client, the configuration typically follows this pattern:

Field	Value
Server name	eConnect (or any name you choose)
URL / Endpoint	https://mcp.econnect.ai/mcp
Transport	http / streamable-http
Authentication	None required at the server level — pass your auth token inside the chat session

Consult your specific tool's documentation for the exact configuration format — the endpoint URL is the only required value.

---

Cursor

Cursor is a popular AI-powered code editor that supports MCP servers.

1. Open Cursor Settings → MCP section
2. Add a new server with the name eConnect, URL https://mcp.econnect.ai/mcp, and type http
3. Save and restart Cursor if prompted
4. In a Composer session (Agent mode), log in with your token:

   Log in to eConnect using this token: [your token]

---

Windsurf (Codeium)

Windsurf by Codeium supports MCP in its Cascade agent.

Add the eConnect server to your ~/.codeium/windsurf/mcp_config.json:

{
  "mcpServers": {
    "eConnect": {
      "serverUrl": "https://mcp.econnect.ai/mcp"
    }
  }
}

Restart Windsurf and use Cascade in agent mode to start querying eConnect data.

---

Custom Agentic Pipelines

For developers building custom agentic workflows with frameworks like LangChain, AutoGen, CrewAI, or direct API orchestration, the eConnect MCP server is accessible via standard MCP client libraries.

Python (using mcp SDK)

from mcp.client.streamable_http import streamablehttp_client

async with streamablehttp_client("https://mcp.econnect.ai/mcp") as (read, write, _):
    # Initialize and use the MCP session
    # Call econnect_get_auth_token, econnect_login, then query tools
    pass

Node.js / TypeScript

import { Client } from "@modelcontextprotocol/sdk/client/index.js";
import { StreamableHTTPClientTransport } from "@modelcontextprotocol/sdk/client/streamableHttp.js";

const transport = new StreamableHTTPClientTransport(
  new URL("https://mcp.econnect.ai/mcp")
);
const client = new Client({ name: "my-agent", version: "1.0.0" });
await client.connect(transport);

Refer to the MCP Protocol documentation for more details.

---

Authentication in Automated Pipelines

For headless pipelines without an interactive chat session, generate a token by calling econnect_get_auth_token at the start of each pipeline run, then immediately call econnect_login:

1. Call econnect_get_auth_token(server, username, password)
2. Call econnect_login(token)
3. Proceed with data queries

Tokens last 7 days by default (up to 30 days, server-configurable) and are reusable — you can pre-generate a token from the Auth Token page, store it securely in your pipeline's secret management system, and reuse it across runs until it expires.

See the Auth Token guide for full details.

Credential Security in Pipelines

Never hard-code eConnect credentials in source code. Use environment variables or a secrets manager to inject credentials at runtime.

---

Supported MCP Protocol Details

Property	Value
Protocol	Model Context Protocol (MCP)
Transport	Streamable HTTP (JSON-RPC over HTTP POST)
Endpoint	POST https://mcp.econnect.ai/mcp
Server version	v1.3.0
Total tools	62
Categories	7
Runtime	.NET 10

For a full list of available MCP tools and their parameters, log in and call econnect_plugins from any connected client.