How Cloudflare is Revolutionizing AI Agent Experiences with Claude and MCP for Leading Tech Companies
Cloudflare, Inc. (NYSE: NET), a global leader in connectivity cloud solutions, has announced a groundbreaking initiative that empowers leading technology companies to deliver seamless, real-world AI agent experiences using Anthropic’s AI assistant, Claude. Through its innovative toolkit powered by Cloudflare Workers, the company enables secure, low-latency connections between AI assistants like Claude and popular SaaS platforms such as Asana, Atlassian, Block, PayPal, Sentry, Stripe, and more. This collaboration marks a pivotal moment in the evolution of AI-driven productivity tools, allowing users to interact with their favorite business applications through natural conversations rather than navigating complex interfaces.
The Challenge of Building Autonomous AI Agents
AI has already transformed how we work, from drafting emails to generating code and analyzing data. However, most AI tools today still require users to switch between multiple applications to execute tasks recommended by the AI. For truly autonomous, agent-based AI experiences—where AI acts on behalf of users—it’s essential for these tools to directly interact with enterprise software. This requires reliable access to external systems and datasets, which presents significant technical challenges, including ensuring security, scalability, and minimal latency on a global scale.
Enter Cloudflare’s Managed Compute Protocol (MCP)—a solution designed to bridge the gap between AI agents and enterprise tools. By providing the first and only simple toolkit to quickly set up remote MCP servers, Cloudflare allows developers to focus on building powerful AI experiences without worrying about the complexities of authentication, authorization, or scaling infrastructure.
As Matthew Prince, co-founder and CEO of Cloudflare, explains, “Cloudflare empowers AI to connect to the world. Whether you’re interacting through a browser, an app, or an AI assistant like Claude, we’re the backbone that makes these experiences fast, secure, and reliable. As agent-based AI becomes the new interface, Cloudflare is the central infrastructure enterprises will rely on to build and scale their AI strategies.”
Simplifying Development with MCP Servers
The open-source Managed Compute Protocol (MCP), developed by Anthropic, establishes a framework for securely connecting AI agents to third-party services. While adopting MCP offers immense potential, implementing it reliably at scale can be technically challenging. Cloudflare addresses this by offering a streamlined toolkit that simplifies the process of setting up remote MCP servers.
With Cloudflare’s toolkit, development teams can deploy secure MCP servers within days instead of weeks. Key benefits include:
- Simplified Authentication: Built-in controls handle complex authentication processes, ensuring secure interactions between AI agents and external tools.
- Granular Permissions: Developers can restrict what actions AI agents can perform, safeguarding sensitive data and operations.
- Real-Time Visibility: Teams gain insights into data access and actions performed, enhancing transparency and accountability.
- Global Scalability: Backed by Cloudflare’s extensive network, these solutions deliver fast, reliable performance regardless of location.
“Building connections reliably at scale is complex,” says Mahesh Murag, Product Manager at Anthropic. “Cloudflare simplifies and secures the process of connecting applications to Claude via MCP, accelerating adoption and fostering an ecosystem of remote server integrations.”
Real-World Use Cases Across Industries
Several industry leaders are already leveraging Cloudflare’s MCP capabilities to enhance user experiences with AI:
Asana
Prashant Pandey, CTO of Asana, highlights how MCP connects their Work Graph directly to AI tools like Claude: “Our integration transforms natural language into structured work—creating projects from meeting notes or feeding updates to AI. Built on Cloudflare’s infrastructure, we deployed quickly and focused on creating the best experience for our users.”
Atlassian
Taroon Mandhana, Head of Product Engineering at Atlassian, emphasizes the importance of integrating AI into existing workflows: “By hosting MCP on Cloudflare, we delivered powerful integrations faster while keeping corporate data secure. From OAuth to out-of-the-box remote MCP support, Cloudflare accelerated our efforts significantly.”
Block & Square
Square’s dedicated MCP server allows sellers to leverage AI agents like Claude to manage payments, inventory, orders, and customer relationships seamlessly. With fewer technical requirements, businesses can unlock sophisticated use cases tailored to their needs.
PayPal
Prakhar Mehrotra, Senior Vice President of Artificial Intelligence at PayPal, calls MCP a “new paradigm for software development.” He adds, “Developers can now delegate tasks to AI agents for seamless integration with PayPal’s commerce features, optimizing workflows autonomously.”
Stripe
Jeff Weinstein, Product Lead at Stripe, predicts MCP could become the standard way people interact with services: “With Stripe’s Agent Toolkit, developers can monetize their MCPs with just a few lines of code.”
Webflow
Utkarsh Sengar, VP of Engineering at Webflow, notes how MCP simplifies web development tasks: “Tasks like managing CMS, improving SEO, and even helping users learn Webflow can now be accomplished by instructing an AI agent—all thanks to Cloudflare’s secure and scalable infrastructure.”
Cloudflare’s Own MCP Servers
In addition to supporting other companies, Cloudflare is launching its own suite of MCP servers. These enable developers to interact with Cloudflare’s platform through Claude, streamlining tasks like analyzing logs, debugging errors, and securing networks—all without navigating documentation or complex tools.
To explore Cloudflare’s growing list of MCP servers, visit GitHub.
