Fixing Security Issues Where They Happen: RAD + IDE via MCP

We’re excited to announce a powerful new capability in RAD Security: developers can now get context-aware, AI-suggested fixes for cloud security issues right inside their IDE, thanks to a new integration between RAD and Cursor.

This integration is powered by the Model Context Protocol (MCP)—an emerging open standard that enables AI tools to communicate with each other using shared context. MCP is still new, but we believe it represents the future of agentic AI: not just generating insights, but acting on them across systems in meaningful ways.

What this means in practice

Let’s say RAD detects a misconfigured manifest or a vulnerable cloud workload. That alert is useful—but for a developer trying to fix it, switching tools, looking up best practices, and writing a secure patch takes time and focus.

Now, agents built in your IDE can use MCP to check with RAD in order to determine whether the open file contains any security issues. Cursor receives the relevant context—affected resources, risk details, even recommended fixes—and uses its in-IDE AI assistant to propose a code-level fix. The developer stays in flow, reviews the suggestion, and can apply it in a click.

Why we’re excited about MCP

What makes this so powerful isn’t just the convenience. It’s the architecture. RAD now acts as an MCP server, speaking a common language that other tools—like Cursor—understand. We’re among the first ~20 security-focused tools in the world to go live with MCP support, and most others are focused on pentesting or offensive security.

This is one of the first examples of a security detection engine handing off rich, contextual data to an AI-enabled IDE—and getting real, actionable results in return.

Where this is headed

Agentic AI is a buzzword right now, but this integration shows what it can actually look like in practice. Not autonomous tools working in silos, but systems that coordinate, collaborate, and take action based on a shared understanding of the world.

We’re just getting started with what’s possible here. In the meantime, we’re thrilled to give security and developer teams a faster, more intuitive way to fix what matters—right where the work happens.

‍