New Relic MCP. Get APM scores and error rates instantly from chat.

Listen up. This server connects your AI agent straight to New Relic via the NerdGraph GraphQL API. Forget clicking through menus or opening a browser—you'll get full visibility into your whole stack using nothing but natural conversation. Your agent handles everything, from figuring out what services you’ve monitored to running raw queries that only an expert would know how to write.

Checking System Health Metrics: You can immediately check the current operational status of any application. When you need a quick look at golden metrics—the error rate, Apdex score, and average response time—you just ask for it. The agent pulls that summary right up so you know if something's going sideways before you even open your laptop.

Auditing Alerts and Incidents: You don’t gotta manually check the alerts dashboard anymore. If you want to track every active issue or high-priority alert across all connected New Relic accounts, you ask for a list. This gives you an immediate rundown of what needs attention right now. Similarly, you can get a full manifest of every configured Service Level Objective (SLO) and see exactly how it's performing against its target goals.

Discovering Monitored Components: Need to know what’s running? You can list all the monitored entities in your account. This includes everything: APM applications, containers, hosts, and even browser sessions. If you have a unique identifier for one of those components, you can fetch detailed metadata on it, giving you a deep dive into that specific piece of infrastructure.

Running Custom Data Queries: When the standard dashboards aren't enough, you run raw New Relic Query Language (NRQL) strings. This lets you pull highly specific metrics and datasets directly from your infrastructure, getting data points nobody else can reach. You just drop in the query, and the agent returns the raw results.

Managing Observability Dashboards: If someone built a custom dashboard you need to look at, you don't have to guess its ID. You can list all available custom observability dashboards within your environment. Need to know what it's configured to show? You fetch the configuration details for any specific dashboard by name or ID.

Understanding Accounts and Users: Your agent provides basic info about the current user and the account connected to the server with get_me. If you need to see which New Relic accounts your API key has access to, you run a simple list of accounts command.

The Process is Simple: You just talk to your AI client. It translates that natural language—whether you're asking for performance data on one container or listing all available dashboards—into specific API calls and sends back clean, structured data instantly.

This server lets you skip the clicks and speak directly to your monitoring backend. You get immediate access to golden metrics summaries, a list of every active alert across accounts, full details on monitored entities like apps and hosts, raw dataset results from custom NRQL queries, and even the configuration details for any dashboard someone built.

It handles all that heavy lifting so you're not stuck in a menu loop. You just tell it what you need to know about your system health or your performance goals. It gives you the data; you do the thinking.