Hadrian's MCP server is live

Starting today, you can connect Hadrian directly to your AI assistant, allowing you to conversationally interact with your external exposure intelligence. With this feature, security teams can query exposures, investigate assets, coordinate fixes across Jira, ServiceNow, Slack, and AWS all through natural language, all without leaving the conversation.

This is the first meaningful shift in how security teams interact with exposure data since the move from spreadsheets to dashboards.

The operational cost of disconnected tools

When an exposure is identified in Hadrian, an engineer gets a notification and opens the dashboard to investigate the finding. They check Jira to see if a ticket already exists for this asset. They switch to AWS to understand the infrastructure context. They move to Slack to coordinate with the team managing that system. They jump to ServiceNow to track the remediation timeline. By the time the problem is routed to the right person with the right context, the engineer has switched contexts five times and lost track of critical details in the process.

This operational friction isn't a minor inconvenience. For teams managing thousands of assets and responding to dozens of exposures daily, context switching becomes the limiting factor in how fast they can respond to risk. The bottleneck isn't identifying problems. It's acting on them.

How it works in practice

Consider the typical workflow for addressing critical exposures. You need to identify which risks require immediate attention, understand what's involved in fixing them, and assign the work to the right team. Today this happens across multiple tools and multiple contexts.

With the MCP server, you start a conversation with your assistant. You ask:

"Group all DNS and email security findings that are assigned to me."

Within seconds, you have a filtered list that includes affected assets, severity, and current status. The next step is understanding what fixing these risks entails. You continue:

"Find a common pattern in all our DNS risks. Identify low hanging fruit."

You now have detailed guidance for each finding, including reproduction steps and recommended fixes. At this point you're ready to assign work. Rather than copying findings into Jira and manually enriching each ticket, you ask:

"Identify the engineering teams responsible for the grouped risks (use Jira MCP), search for matching tickets and create tickets for each engineering team if they weren't raised before."

Tickets appear in Jira with full context attached. Not empty tickets requiring manual enrichment. Tickets that already contain affected assets, remediation guidance, and severity information. Your team receives work-ready assignments they can begin addressing immediately.

In the traditional workflow, each step happens in a different system. Filtering and investigation in Hadrian. Ticket creation and assignment in Jira. Context reconstruction at every handoff. Here, the entire process flows as a single conversation. You filter, investigate, and assign without breaking context or switching tools.

Investigating the attack surface in depth

Your security team often needs to answer questions that require correlating information across multiple data sources. Consider asking your assistant:

"List all production domains that are publicly exposed and do not have a WAF."

Your assistant returns it in seconds. Once you have the initial list, investigation becomes natural:

"Which of these domains have open ports 22 or 3389?"

The response narrows to assets with potentially dangerous exposure. You might then ask:

"Where are we running unpatched versions of OpenSSL?"

You get back a complete picture of your vulnerable software landscape across the assets you're investigating. These investigations normally require combining multiple filters and views within Hadrian, then cross-referencing with cloud platform data, then manually tracking relationships between domains, IP addresses, services, and technologies. Your assistant assembles all of this context and returns it in a single response. An investigation that would take a security engineer 30 minutes using the dashboard takes seconds through conversation.

Reporting and strategic visibility

Security teams need to communicate posture not just to their peers but to leadership, boards, and external stakeholders. Generating these reports today requires exporting data from Hadrian, compiling information from multiple parts of the platform, and formatting it for presentation.

With the MCP server, you ask your assistant:

"Generate a security summary for this week, including new risks by severity and current score."

You receive a breakdown of newly discovered risks, severity distribution, and your current security score. When you need to understand trends:

"Show how our critical risk count has changed over the past 90 days."

A time-based view appears showing whether your exposure is increasing or decreasing over time. These summaries are immediately usable in board presentations, stakeholder updates, and internal communications. They're generated on demand using the same underlying data your team uses for daily operations, which means they're always accurate and current. No exporting. No manual compilation. No delay between when data changes and when leadership understands it.

Coordinating across your entire security stack

The full power of the MCP server emerges when your assistant can see multiple systems simultaneously. If you've connected Hadrian alongside Jira, ServiceNow, AWS, and Slack, your assistant becomes a coordination engine for your entire security operations workflow.

You can ask:

"For every new critical risk discovered this week, create a Jira ticket with the affected assets and remediation steps."

Your assistant identifies the risks in Hadrian and creates tickets in Jira with full context attached. Ticket creation that would normally take hours if done manually happens instantly, at scale. You can correlate data across systems just as easily:

"Check which of these assets are hosted on AWS and have open ports."

Hadrian data combines with AWS configuration data to show you exactly what's exposed in your cloud infrastructure. You can notify your team directly:

"Post a summary of new critical risks in the last 24 hours to the security Slack channel."

The assistant generates the summary and posts it automatically. What previously required bouncing between tools, waiting for information to be gathered, and coordinating handoffs between teams now happens in a single interaction. Work that would consume hours of your team's time becomes automatic.

Getting started

The Hadrian MCP server is available in preview as of today. The server endpoint is https://mcp.hadrian.io.

Connect it to any MCP-compatible AI assistant. Whether you're using Claude Desktop, another LLM platform, or a custom agent, the setup process is similar: add the server URL to your MCP configuration and authenticate with your Hadrian credentials.

See the documentation for setup instructions specific to your platform: https://docs.hadrian.io/integrations/mcp

Once connected, your assistant has access to all 50 Hadrian tools across risk management, asset intelligence, vulnerability data, and write operations. If you've also connected other MCP servers for Jira, ServiceNow, AWS, and Slack, your entire security stack becomes available within a single conversation.

Start with simple queries to understand what's available. Ask your assistant about your current security grade or create critical risk tickets for any exposure you haven't tracked yet. Your assistant learns your operational context quickly.

What this means for security teams

Your team has always had the intelligence they need to manage risk. They understand what's exposed. They know what needs fixing. The constraint has never been knowledge. It's been the operational overhead of acting on that knowledge across disconnected systems.

The MCP server eliminates that overhead. It doesn't change what you know. It changes how fast you can act on what you know. Your team moves from spending time on tool coordination and context reconstruction to spending time on actual security work. Risk moves from discovery in Hadrian to assigned work in Jira to remediation tracking in ServiceNow without anyone on your team having to manually bridge those systems.

Connect Hadrian to your AI assistant today and see the difference immediately.