What can a widget chart?

Any gauge, sum or histogram metric the project has reported, from the AdonisJS agent or from the OpenTelemetry metrics API. Saving a widget checks that every metric name exists in the project.

Who can change a dashboard?

Its creator and org admins. Everyone who can view the project can open it read-only and duplicate it into a copy of their own. Org admins manage dashboards that have no creator on record.

How many widgets fit on a dashboard?

Up to 60, on a 12-column grid, each at most 12 rows tall. Edit layout lets you drag, resize, duplicate and remove widgets, and Save stores every change at once.

Will the person I send a link to see the same data?

Copy link keeps the time range, environment, services and variable values, but a relative range such as the last hour ends when the link is opened. Pick Copy link to this exact window to pin the start and end.

What happens when a widget cannot load?

It shows the reason in its body instead of an empty chart, such as a metric the project never reported or a Bars widget with a single series. When telemetry storage is unavailable, the dashboard still opens and each widget offers Retry.

Dashboards

Put the metrics you watch on one page.

Chart request latency next to heap usage next to CPU, from the OpenTelemetry metrics your services already report. Variables filter every widget at once, and a copied link opens the same view for anyone on the project. Every project also opens on an overview of its health.

Six visualizations: Time series, stats, bars, pies and tables
Variables: One selection filters every widget
Formulas: Combine queries, such as A / B
Shared by link: Same range, services and variable values

#variables

Filter every widget at once

A new dashboard comes with $service, $env, $version and $route. Pick values in a chip and every widget whose metric carries that attribute narrows to them.

Up to 8 variables, each tied to one attribute your metrics carry
A widget filter can use $service as a value, and drops it when nothing is selected
Selections live in the URL as var.service=api, so a link keeps them

#widgets

Build a widget against live data

The widget editor previews the chart with the current time range and services while you pick a metric, an aggregation, a group by and filters. Aggregations follow the metric type, so a histogram starts on P95 and a counter on rate per second.

Time series, stat, sparkline, bars, pie and table visualizations
Up to 6 queries per widget, each drawing its 10 largest groups
Drag and resize on a 12-column grid, or move widgets with the arrow keys

#formulas

Ratios and percentages from formulas

Formulas combine a widget's queries by letter into one series, such as A / B for a ratio or A * 100 for a percentage. An assistant can build the same widget through the MCP server.

Up to 4 formulas per widget, each with its own legend
Numbers, +, -, *, / and parentheses between query letters
Values format as numbers, durations, percentages or bytes, or follow the metric's unit

add_widget.json

{
  "project": "storefront",
  "dashboardId": 2,
  "title": "Heap used of limit",
  "type": "chart",
  "queries": [
    {
      "metric": "v8js.memory.heap.used",
      "aggregation": "max",
      "filters": [{ "key": "service.name", "values": ["$service"] }]
    },
    {
      "metric": "v8js.memory.heap.limit",
      "aggregation": "max",
      "filters": [{ "key": "service.name", "values": ["$service"] }]
    }
  ],
  "display": {
    "values": "percent",
    "formulas": [{ "expression": "A / B", "legend": "heap used" }],
    "hideQueries": true
  }
}

#add-to-dashboard

Keep a chart you found in Metrics

Add to dashboard on a metric's page turns the chart into a time series widget with the same aggregation, grouping and filters. Pick an existing dashboard or name a new one.

The widget keeps the chart style, line, area or bars
Filters other than "is" are listed as not carried over
A new dashboard and its first widget are saved together

#overview

An overview for every project

Each project opens on an overview that needs no setup. It reads the traces your services send and tells whoever is on call whether the app is healthy and what to look at first.

Alerts, uptime, incidents and new issues in one strip, refreshed every 30 seconds
Eight cards with a sparkline and the change against the previous period
Traffic and latency charts with deployment markers, then the slowest endpoints and queries

Works with

Where the numbers come from

#faq

Questions

In the docs

How it works, in detail

Every feature

Try it on your own stack.

Every feature is on every plan, priced by the telemetry you send.