Announcing: Native support for MCP Servers

Announcing: Native support for MCP Servers

Nate Barbettini's avatar
Nate Barbettini
MARCH 28, 2025
2 MIN READ
MCP
Rays decoration image
Ghost Icon

What is MCP?

Model Context Protocol (MCP) is having a moment, and for good reason: it's an open, standard way to connect tools and agents. At Arcade, we're thrilled to see the industry adopting an open standard for tool calling, because it means even more options and flexibility for devs building agents.

Easily connect to any MCP Server with Arcade

Today, we're excited to release native support for MCP servers in Arcade! Arcade can now connect to any MCP server supporting the new streamable HTTP transport. This means you can seamlessly combine MCP tools and Arcade's tools in your agent or AI app.

What is MCP’s Streamable HTTP transport? 

MCP’s streamable HTTP transport is hot off the press—just finalized 3 days ago. It is a simple way for apps to talk to AI tools over the internet.

It works like this: your app sends a message to an MCP server, and the server can either reply once immediately or keep a connection open to send multiple messages over time.

Why it matters:

  • Makes it easy to connect AI tools together
  • Supports both quick exchanges and longer conversations
  • Handles connection drops gracefully

Now that MCP servers can operate over the network (as opposed to only on your local machine), MCP can help developers build agents and AI apps that run in the cloud. Instead of every company building their own closed system, MCP lets developers mix and match tools from different providers - making AI more flexible and accessible for everyone.

Benefits of using MCP with Arcade

MCP's streamable HTTP transport creates an open standard for connecting AI tools, and Arcade's new native support makes this even more powerful for developers.

With MCP servers in Arcade, you get:

  • Arcade's prebuilt tools with deep integrations (Slack, Gmail, X, Salesforce)
  • Easy access to build your own custom tools on top of Arcade
  • Access to any tool built by others following the MCP standard
  • Arcade's authorization management to handle the security complexity

The key benefit is flexibility: developers can now freely mix and match tools from different providers without custom integration work. This flexibility helps developers create more powerful AI applications by combining specialized tools from across the ecosystem exactly as needed for your specific use case.

How to get started with an HTTP MCP Server

We expect MCP’s streamable HTTP transport will become widely implemented in servers soon. Until then, we've created a simple HTTP MCP server for you to experiment with at https://mcp-http-demo.arcade.dev.

  • Log in to your Arcade dashboard (or create an account if you haven't already) to try it out. 
  • Go to Playground > Execute and try executing the lorem tool. 
  • MCP achievement unlocked! 🏅
  • Check the Workers section to see the mcp-demo server in the workers list. 
  • Add another HTTP-compatible MCP server by clicking Add worker > MCP

How to handle authorization for MCP

Also hot off the press is OAuth 2.1 authorization for MCP, which we are testing in private beta. Securing MCP servers correctly is easier said than done—stay tuned for an upcoming deep-dive post on doing it the right way. 

Interested in joining our beta? Reach out via email or Discord.

What's next?

We're working on a comprehensive deep-dive on implementing secure MCP servers. More integration capabilities are coming soon to make your development experience even smoother.

Follow along by signing up at www.arcade.dev

Happy integrating!

SHARE THIS POST

RECENT ARTICLES

THOUGHT LEADERSHIP

The Agent Hierarchy of Needs: Why Your AI Can't Actually Do Anything (Yet)

Your AI can summarize documents you feed it, answer questions about your uploaded PDFs, and explain concepts from its training data. But ask it to pull your actual Q4 revenue from NetSuite, check real customer satisfaction scores, or update a deal in Salesforce? Suddenly it's just guessing—or worse, hallucinating numbers that sound plausible but aren't your data. This disconnect between AI's intelligence and its ability to access real data and take action is why less than 30% of AI projects hav

COMPANY NEWS

We Just Won "Overall Authentication Solution of the Year" — Here's Why It Matters for AI Builders

Arcade.dev just took home "Overall Authentication Solution of the Year" in the 8th Annual AI Breakthrough Awards. And before you roll your eyes at another tech award announcement, let me explain why this actually matters for anyone building AI agents that need to do real work. The Problem We All Keep Hitting You know that moment when your perfectly crafted AI agent suggests "I'll schedule that meeting for you" — and then... doesn't? Because it can't? Yeah, that's the wall everyone's been hit

Rays decoration image
TUTORIALS

How to choose the best Agentic Framework, Part 2: Agentic Delegation

In the previous post in this series, we explored Human-in-the-Loop. Here, we’re exploring Handoffs, which I prefer to call “Agentic Delegation” This post is a companion to a video, I encourage you to watch it! Here’s the experiment setup I’m using the same agentic system. I implemented the same system using three different Frameworks: * LangGraph * OpenAI’s Agents SDK * Google’s Agent Development Kit (ADK) In all cases, the agent uses a “supervisor” architecture, where a single agent re

Blog CTA Icon

Get early access to Arcade, and start building now.