Create Dashboard with Kopai

Build a dashboard by writing a short TypeScript script: design a uiTree, feed each data tile with a type-safe kq query, and create it via client.createDashboard(...). The CLI stays as a fallback for one-off creates.

Component Schema (auto-generated)

!npx @kopai/cli dashboards schema 2>/dev/null || echo "ERROR: Cannot connect to Kopai backend. If running locally, start it with: npx @kopai/app start — If using a remote backend, check the url in your .kopairc file."

Available Metrics

!npx @kopai/cli metrics discover --json 2>/dev/null || echo "ERROR: Cannot connect to Kopai backend. If running locally, start it with: npx @kopai/app start — If using a remote backend, check the url in your .kopairc file."

Code mode (recommended)

Connect — clientFromConfig() reads .kopairc just like the CLI:

import { clientFromConfig } from "@kopai/sdk/node";
const client = clientFromConfig(); // ./.kopairc → ~/.kopairc → http://localhost:8000

Run — npx tsx dashboard.mts. Use .mts so top-level await works regardless of the project's module type. Requires @kopai/sdk installed (npm i @kopai/sdk).

Output — console.log(dash.id) so the new dashboard id lands on stdout. Wrap the build + create in try/catch.

Typed results — discoverMetrics() and client.query(q) return fully typed values (metric name/type/unit, aggregate measure values as number, etc.). When you probe/verify data, iterate the results directly — don't cast to any/any[].

Two rules that will save you a broken dashboard

1. Feed data tiles with a RAW query — kq.metrics("<Type>").raw()…build(), not .aggregate().

The tile renderers (MetricStat/MetricTimeSeries/MetricHistogram/MetricTable, LogTimeline, TraceDetail) plot raw rows; an aggregate/summary/timeSeries query is rejected at render with "this panel displays raw metric rows … use a raw metric query". Choose the MetricType up front via kq.metrics("Gauge") (auto-pinned), add a MetricName filter and a time window on metric tiles.

1. Set every declared prop — use null for unused ones. The render-time schema

requires nullable props to be present, not omitted. Omitting them passes createDashboard but fails at render with "invalid layout: …props.description: expected string, received undefined". So a Card is { title, description: null, padding: null }, a MetricTimeSeries is { height, showBrush: null, yAxisLabel: null, unit }, etc.

Backend caveats

• Every tile query needs a time window — .timeRelative("1h") / .timeAbsolute(startISO, endISO).
• Choose the MetricType up front via kq.metrics("Gauge") — it's auto-pinned, no manual .where("MetricType", …) needed. Use the type from discoverMetrics(). Types: Gauge | Sum | Histogram | ExponentialHistogram | Summary.
• The SDK field is uiTreeVersion (the CLI flag is --tree-version). Current tree version: 0.14.0.

Workflow

1. Discover metrics — const { metrics } = await client.discoverMetrics(); Each entry

is { name, type, unit, description, attributes, resourceAttributes }. Pass the metric's type to kq.metrics("<Type>"), and set the tile unit prop from unit.

1. Build each tile's query with kq.metrics("<Type>").raw() (or kq.logs.raw() /

kq.traces.raw()) — returns a KopaiQuery object; hold it in a variable.

1. Assemble the uiTree — embed the built query into the tile:

dataSource: { method: "query", params: <builtQuery> }. Set every prop (null for unused).

1. Create —

``ts const dash = await client.createDashboard({ name: "<name>", uiTreeVersion: "0.14.0", uiTree, metadata: {}, }); console.log(dash.id); ` On a KopaiValidationError/KopaiError, re-run discoverMetrics()` to recheck names and types, fix the tree, and retry.

1. View — open <baseUrl>/?tab=metrics&dashboardId=<dash.id>.

Quick Example

// create-cpu-dashboard.mts — run: npx tsx create-cpu-dashboard.mts
import { kq } from "@kopai/sdk";
import { clientFromConfig } from "@kopai/sdk/node";
const client = clientFromConfig();

// Tile data: a RAW metric query (tiles render raw rows, not aggregates).
const cpuSeries = kq
  .metrics("Gauge") // MetricType chosen up front, auto-pinned
  .raw()
  .where((f) => f.eq("MetricName", "system.cpu.utilization"))
  .timeRelative("1h")
  .limit(500)
  .build();

const uiTree = {
  root: "stack-1",
  elements: {
    "stack-1": {
      key: "stack-1",
      type: "Stack",
      props: { direction: "vertical", gap: "md", align: null },
      children: ["card-1"],
      parentKey: "",
    },
    "card-1": {
      key: "card-1",
      type: "Card",
      props: { title: "CPU Usage", description: null, padding: null },
      children: ["ts-1"],
      parentKey: "stack-1",
    },
    "ts-1": {
      key: "ts-1",
      type: "MetricTimeSeries",
      props: { height: 300, showBrush: null, yAxisLabel: null, unit: "1" },
      children: [],
      parentKey: "card-1",
      dataSource: { method: "query", params: cpuSeries },
    },
  },
};

const dash = await client.createDashboard({
  name: "CPU Dashboard",
  uiTreeVersion: "0.14.0",
  uiTree,
  metadata: {},
});
console.log(`Created dashboard ${dash.id}`);

Components

Layout (have children, no dataSource): Stack, Grid, Card. Static: Heading, Text, Badge, Divider, Empty.

Data tiles — feed each with a raw kq query via method: "query" (legacy methods still work, listed for fallback):

Sizing/props: set height: 300 on MetricTimeSeries/MetricHistogram, height: 600 on LogTimeline (smaller collapses it to a count badge); MetricStat needs no height. Set unit on chart tiles to the metric's raw OTEL unit ("By", "s", "ms", "1"). And remember rule #2: every declared prop must be present (null for unused).

Rules

• workflow - Dashboard creation workflow (tree structure, dataSource rules, layout, error handling)

Read rules/<rule-name>.md for details.

CLI fallback (one-off create)

The CLI takes the tree on stdin and uses --tree-version (the SDK field is uiTreeVersion). It cannot build kq queries, so tiles use the legacy searchMetricsPage/searchLogsPage form:

echo '{"uiTree":{…},"metadata":{}}' | npx @kopai/cli dashboards create --name "<name>" --tree-version "0.14.0" --json