New Relic
Query New Relic with NRQL, search entities, manage tags and alerts, read dashboards and record deployments — all through NerdGraph.
About
New Relic ships in the w6w first-party pack. It declares 17 actions, 3 health checks, and the host runs its code in a sandbox that never sees the credential.
- App id
io.w6w.newrelic- Version
- 0.1.1
- Author
- w6w
- Licence
- MIT
- Categories
- Monitoring & Observability · Analytics & BI · DevOps & Infrastructure
Overview
New Relic exposes almost everything through one GraphQL endpoint, NerdGraph, and this app puts the parts a workflow actually needs on top of it: query data with NRQL directly, or fall back to a raw GraphQL query for anything not wrapped; search entities and read them in detail; attach or remove tags; list alert policies and conditions; read and acknowledge open issues; mark a deployment on the charts; read dashboards down to the NRQL behind each widget; and list synthetic monitors.
Because NerdGraph reports failures inside an HTTP 200 rather than as an error status — sometimes in a top-level errors array, sometimes buried inside the mutation’s own response — this app treats a plausible-looking but incomplete result as a failure rather than a silent partial success, and every action that performs a mutation confirms something actually changed. NRQL queries default to the last hour and the first 100 rows when not told otherwise, which the action names explicitly rather than leaving as a surprise; entity search reports entities that stopped reporting data rather than dropping them; and alert conditions are checked for ones that are disabled or would not fire on missing data.
New Relic accounts live in exactly one of two regions, each with its own data and its own entity ids that never resolve in the other, and a single ambiguous “authentication required” error can mean a wrong key, the wrong key type, or simply the wrong region — this app distinguishes all three rather than leaving a workflow to guess.
Build with New Relic
Three routes to the same 17 actions. The Workflow tab is generated from New Relic's own manifest and carries its real ids, so it is copy-pasteable; the Code and CLI examples are the same call for any action on any app, so every app-specific value in them is a blank you fill in.
List accounts
account-list Accounts this key can reach. An account in the OTHER region is absent from this list entirely rather than listed and empty.
List alert conditions
alert-condition-list The NRQL conditions in a policy. Reports how many are DISABLED, and how many would stay silent if data stopped arriving entirely.
List alert policies
alert-policy-list Alert policies on an account. A policy with no conditions is legal and watches nothing — `alert-condition-list` is what answers whether anything is actually monitored.
Get a dashboard
dashboard-get A dashboard with its pages and widgets — including the NRQL behind each, which is the practical way to lift a query out of a dashboard and run it elsewhere.
List dashboards
dashboard-list Dashboards on an account. There is no dashboards endpoint — a dashboard is an entity, found through entity search like everything else.
Record a deployment
deployment-create Mark a deployment on an entity's charts — the thing that makes 'did anything ship' answerable. The timestamp must be within 24 hours of now, either direction.
Get an entity
entity-get One entity in detail. `alertSeverity` of NOT_CONFIGURED means nothing is watching it at all — a different condition from healthy, and easy to read as one.
Search entities
entity-search Find applications, hosts, monitors and dashboards, and get their GUIDs. Entities that stopped reporting stay searchable for days — those are counted separately.
Add tags to an entity
entity-tag-add Attach tags, which is how alerts, dashboards and workloads select things. Adding to an existing key APPENDS rather than replacing.
Remove tags from an entity
entity-tag-delete Remove a whole tag key, or specific values. Alerts and workloads SELECT on tags, so removing one can quietly drop an entity out of what was watching it.
Run a NerdGraph query
graphql-query Send an arbitrary GraphQL query or mutation. The HTTP and GraphQL error levels are still checked; a raw MUTATION should request `errors { message type }` and read them itself.
Acknowledge an issue
issue-acknowledge Move an issue to ACTIVATED so everyone can see it is being handled. It does not stop the condition evaluating and does not close anything.
Close an issue
issue-close Resolve an issue. The condition keeps evaluating — closing something still breaching just opens a new issue on the next incident.
List issues
issue-list What is currently wrong. An ISSUE groups incidents and is what a person is paged about — filtering to CREATED alone omits everything already being worked on.
Run an NRQL query
nrql-query Query New Relic's data with NRQL. Note the two silent defaults: no SINCE means the last HOUR, and no LIMIT means the first 100 rows.
List Synthetics monitors
synthetics-monitor-list Uptime monitors on an account. A disabled monitor is silent rather than failing, so 'no alerts' and 'nothing running' look identical — both are counted here.
Get the current user
user-get Who this key belongs to. A user key carries exactly that person's permissions, so this is also the answer to what the connection can do.
A workflow step names the app and the action, and the editor fills in the
connection when you pick one. This is the Step shape from the
workflow spec, carrying New Relic's real ids.
{
"manifestVersion": "2",
"name": "newrelic-example",
"steps": [
{
"id": "dashboard-get",
"uses": {
"app": "io.w6w.newrelic",
"action": "dashboard-get",
"connection": "conn_YOUR_CONNECTION_ID"
},
"with": {
"guid": "<guid>"
}
}
]
}Here are some of the things you can do
-
Get a dashboard
readdashboard-get -
List accounts
readaccount-list -
List alert conditions
readalert-condition-list -
List alert policies
readalert-policy-list -
List dashboards
readdashboard-list
+12 more actions available
Every app-specific value here is a blank you have to fill in. An
app action is reached through the connection that authenticates it, so the
address is a connection id, not the app id — and connections belong to your account,
so a public page cannot know yours. Create one for New Relic, then fill in
the three blanks: conn_YOUR_CONNECTION_ID, the action key, and the
parameters that action declares. The call itself is real — the shape is transcribed
from the studio's own snippet builder, which prints the same kind of blanks — but
nothing in it is specific to New Relic. The Workflow tab is where this app's
real ids are.
npm install @w6w/sdkyarn add @w6w/sdkpnpm add @w6w/sdkdeno add npm:@w6w/sdkimport { W6wClient, isActionRun } from "@w6w/sdk";
// Reads W6W_BASE_URL and W6W_TOKEN from the environment when omitted.
const client = new W6wClient();
const envelope = await client.run({
urn: "conn_YOUR_CONNECTION_ID",
action: "dashboard-get",
payload: {
guid: "<value>",
},
});
if (isActionRun(envelope)) console.log(envelope.value);npm install -g @w6w/cliw6w run conn_YOUR_CONNECTION_ID --action dashboard-get --payload '{"guid":"<value>"}'Give an AI agent New Relic — without giving it New Relic's credentials. One MCP endpoint exposes every app, function and workflow the caller is entitled to, as tools it can discover and run. Access is granted per team while we onboard.
One tool call{
"name": "w6w_invoke",
"arguments": {
"ref": "app:io.w6w.newrelic#dashboard-get",
"input": {
"guid": "<guid>"
}
}
}
Every tool names its target with a single ref. The
app: form above doesn't name a connection at all — the
host resolves which of the caller's New Relic connections to sign
with, and refuses rather than guesses when the answer is ambiguous.
What the agent gets
Credentials it can't read
The token is attached host-side, at the moment of the call. It is never a tool argument, never in the model's context, and never in a transcript — so a prompt injection has nothing to exfiltrate.
A tool surface scoped to the caller
Tools are derived per end user from what that person has actually connected and is entitled to — not one shared bot identity carrying the union of everyone's access.
A durable workflow in one call
Multi-step work runs on the workflow engine and returns a run handle the agent can poll — retries, branching and state survive the conversation that started them.
Health-aware discovery
New Relic's declared health checks are on the surface too, so an agent can tell "the vendor is down" from "your credential expired" before it burns a retry on either.
The MCP surface is part of the hosted platform. New Relic itself is MIT, and the runtime that executes it is source-available (FSL).
Health checks
New Relic declares its own checks, so its health is a property of the app rather than something the host guesses at.
New Relic service status
New Relic's status page, which is partitioned by data centre. An incident in another region is not an incident for an account in this one, so the affected regions are named.
Entities reporting
What proportion of this account's entities have stopped sending data. Monitoring that has gone quiet looks identical to everything being fine — and silences the alerts that would have caught it.