Introducing Coralogix’s MCP Server: Helping customers build smarter AI agents

Now available: Secure, real-time access to your observability data via Coralogix’s Model Context Protocol (MCP) Server.

AI agents are only as powerful as the context they’re given. Today, we’re excited to announce the launch of the Coralogix MCP Server, which enables third-party AI agents to connect directly to your observability data across production, staging, and other environments.

With the MCP Server, customers can enhance their AI agents with access to logs, metrics, traces, SIEM data, and real user monitoring (RUM) straight from the Coralogix platform. This dramatically reduces mean time to resolution (MTTR), streamlines agent workflows, and cuts engineering overhead.

What is an MCP server?

An MCP server is an open standard developed by Anthropic, creators of Claude, that allows tools, data, and services to connect easily with AI models and systems.

Coralogix’s MCP Server acts as a secure gateway between your telemetry and any AI agent you’re building or using. It gives agents access to rich, detailed information about your applications and infrastructure. As agents consume this data, they become more effective and more aligned with your environment.

From Olly to open access

Earlier this year, we introduced Olly, our agentic AI observability assistant that can analyze production systems, understand the context of logs, metrics, and traces, and surface both root cause and business impact.

The MCP Server takes that same deep observability context and opens it up to your own agentic tools. Whether you’re building internal copilots, integrating with an AI-powered IDE, or enhancing a chatbot, you can now:

• Stream live telemetry into any agent
• Connect directly to Coralogix observability data
• Shape the agent experience to match your workflows

Why this matters

AI agents typically operate with limited awareness of system-specific telemetry. Without access to observability data, they struggle to deliver real value in engineering workflows.

What sets Coralogix’s MCP Server apart is its ability to expose observability data that reflects the unique setup of each customer. It can find and retrieve custom attributes, application-specific entities, and context that general-purpose tools can’t. This results in smarter, more accurate agents that understand exactly what’s happening in your systems.

And because the MCP Server supports natural language queries, your agents can locate key events, detect anomalies, and explain metrics without needing hand-coded rules or queries.

Use it where you work

The MCP Server integrates with the tools developers already rely on, such as the popular AI code editor Cursor or your existing IDE.

This integration allows agents to go beyond passive alerts. They can actively detect issues, diagnose root causes, and help resolve them within the same interface, closing the loop on observability and action without tool switching.

Built for AI builders

“Adding the MCP server to our current AI capabilities will enable teams to create custom AI-driven observability experiences,” said Liran Hason, VP of AI at Coralogix. “Now, our customers can easily equip their AI agents with direct access to production observability data. Publishing an official MCP Server also allows our customers to rely on a trusted MCP source and ensure they get the best and most reliable observability capabilities for their agents.”

Whether you’re building internal tooling, enhancing developer productivity, or bringing observability into your AI workflows, the Coralogix MCP Server makes it easy to connect the dots. It’s available today for all Coralogix customers.

Get started

If you’re already using Coralogix for observability, you’re ready to begin. The MCP Server is available now and can be configured to power any AI agent with secure, real-time access to your telemetry.

Visit docs to learn more or reach out to your Coralogix contact to get started.