Full read-write Claude connector for New Relic

Let Claude query telemetry and manage alerts and incidents in New Relic

Stop tab-switching to investigate and triage. Claude runs NRQL queries, tunes alert conditions, and acknowledges incidents right where your observability data already lives, applying the gated changes only once you approve.

Build my connector Talk to us

New Relic is an observability platform where teams watch their applications, infrastructure, and logs, run NRQL queries, and respond to incidents. People want Claude to do more than read dashboards: to query telemetry, manage alert policies, and acknowledge incidents from chat.

What is a New Relic MCP server?

A New Relic MCP server is a small hosted service that exposes New Relic's data and actions to Claude over the Model Context Protocol, so Claude can issues, code, and incidents inside your own account. We build it as a custom connector with separate read and write tools, and destructive actions stay gated behind your approval.

What can New Relic's current connector do today?

Today, New Relic offers its own MCP server and read-leaning Claude integrations, so Claude can run NRQL queries and surface entity and incident data. Write actions such as creating alert conditions, acknowledging incidents, or editing dashboards are typically limited. A custom New Relic connector adds those writes with scoped, human-in-the-loop controls.

What does full read-write New Relic access unlock?

Full read-write New Relic access lets Claude query telemetry with NRQL, create or tune alert conditions and policies, acknowledge or close incidents, add tags to entities, and update dashboards, all once you approve. The investigation and triage that used to mean tab-switching happen where your observability data already lives.

Run a NRQL query and summarize the result for a service or account
Create or update an alert condition or notification policy
Acknowledge or close (resolve) an open incident
Add or remove tags on an entity to organize your inventory
Update a dashboard widget or NRQL query after you approve it
Mute an alert with a muting rule during a planned maintenance window

Can Claude run NRQL queries in New Relic?

Yes. Claude can run a NRQL query and summarize the result for a service or account, then look up entity and incident data through NerdGraph. These reads run with the scope of the New Relic API key you provide, so Claude sees exactly the telemetry your own key already has access to.

Can Claude create alert conditions in New Relic?

Claude can create or update alert conditions and notification policies, add a muting rule for a maintenance window, and tag entities to organize your inventory. These are separated write tools through NerdGraph, scoped to your key. Destructive changes like deleting a condition wait for your confirmation before anything is applied.

Can Claude acknowledge or close incidents in New Relic?

Claude can pull incident context and propose a response, then acknowledge, unacknowledge, or close (resolve) an open incident. These are gated write actions: Claude prepares the change, but it only runs after you approve it, so nothing about an incident shifts in your account by surprise.

How a New Relic connector works

We build a small MCP server that connects to New Relic with scoped API keys through NerdGraph, using only the permissions the work needs. Read tools (NRQL query, entity lookup) and write tools (alerts, incidents, dashboards) are separated, and destructive or outbound actions like closing an incident or deleting a condition require your confirmation. Claude only ever has the access your own New Relic key already has, and the server can be IP allowlisted if your account requires it.

Frequently asked questions

Can Claude create alert conditions in New Relic, not just read NRQL?

Yes. A custom New Relic connector exposes write tools that create and update alert conditions and policies through NerdGraph, with the key's scope. Many default integrations lean read-only or focus on NRQL queries, which is why managing alerts from Claude usually needs a purpose-built connector with separated, confirmable write actions.

Will Claude close or acknowledge an incident without my approval?

No. Acknowledging or closing (resolving) an incident is a separate, gated write action. Claude can pull the incident context and propose a response, but the change is only applied after you approve it. You decide which actions run automatically and which require a confirmation step first.

Does the connector get more access than my New Relic key has?

No. The connector authenticates with the scoped New Relic API key you provide and inherits exactly its permissions and account access through NerdGraph. It never grants Claude more access than the key already has, and you can rotate or revoke that key at any time to cut access.

Related connectors

Bitbucket connector CircleCI connector Jenkins connector

Tell us what you run. We'll build the connector.

Pick your tool, choose what Claude should do, and get an instant scope and price. No call required.

Build my connector Talk to us