New Relic stores your application performance data, error rates, distributed traces, and alert configurations. When something goes wrong, the workflow is: get alert → open one.newrelic.com → find the right entity → run an NRQL query → look at traces → correlate with recent deploys. Each step is navigation you've done hundreds of times.
An MCP server for New Relic lets your AI agent handle the lookups. Ask about error rates, run NRQL queries, check alert status, or review transaction performance without opening a browser tab.
Once your AI agent has New Relic access:
The AI agent gets whatever access your browser session has. Scope using role-based access within New Relic.
Backend developers: Check error rates while coding. Ask about slow transactions before and after your change. No context switch to diagnose performance issues.
SRE/On-call engineers: Triage alerts faster. Ask "what's happening?" and get NRQL-backed answers inline. Correlate alerts with recent deployments.
DevOps engineers: Monitor deployment health across services. Compare metrics before/after releases without building custom dashboards.
Team leads: Get quick health checks across services. Ask about SLA compliance without navigating through multiple New Relic views.
| New Relic Dashboard | AI Agent |
|---|---|
| APM > select service > error rate chart | "What's the error rate for api-prod?" |
| Query builder > write NRQL > execute | "Run NRQL: SELECT average(duration) FROM Transaction SINCE 1h ago" |
| Alerts > open incidents > filter by priority | "Any critical alerts right now?" |
| APM > Transactions > sort by duration | "What are the slowest endpoints?" |
Both work. The AI agent is faster when you're mid-incident and need answers without navigating multiple views.
New Relic pairs well with other monitoring tools:
Open one.newrelic.com and browse your account normally. Check APM dashboards, run NRQL queries, view alerts. DataFaucet captures everything as callable tools. Deploy, connect to your editor, start querying your observability data from wherever you're coding.
Create your New Relic MCP server in 60 seconds.
Try with New Relic →Point DataFaucet at New Relic and get a working server in 60 seconds.
Create New Relic server free →After creating, add to Claude Desktop:
"new-relic": {
"url": "https://datafaucet.dev/api/mcp/YOUR_ID/sse"
}Give AI agents read access to New Relic. Query APM metrics, check error rates, and browse dashboards from your editor.
Turn Dynatrace into an MCP server. AI agents can check service health, query metrics, and investigate problems from Claude, Cursor, or Windsurf.
Give AI tools access to Datadog metrics, logs, and APM traces. Monitor infrastructure, investigate incidents, and query observability data.
See how DataFaucet compares
Point at any URL. Get a working MCP server in 60 seconds. No API docs needed.
Get notified when new integrations launch
New MCP server guides and templates every week.