Migrating from Datadog

KubeSense can recreate your existing Datadog dashboards as native KubeSense dashboards using the AI dashboard assistant (the KubeSense dashboard skill) together with the MCP Server. The assistant reads an exported Datadog dashboard, discovers the matching metric and label names from your live telemetry, maps each widget to the equivalent KubeSense panel, and produces an importable dashboard JSON that you review before anything is created.

Overview

Migration is a guided, three-step flow:

Export the dashboard JSON from Datadog.
Generate the equivalent KubeSense dashboard by asking the AI assistant — it uses the MCP server to validate metric and label names against your real data, so each panel actually renders.
Import and review the generated JSON on the Dashboards page. Nothing is created until you confirm.

Because the assistant discovers names from your live data (Datadog and KubeSense rarely name metrics and labels identically) and you review the result before saving, you get a working KubeSense dashboard rather than a literal, possibly-empty copy.

Prerequisites

Install the dashboard skill in your AI client (Claude Code, Codex, Cursor):

Step 1 — Export your Datadog dashboard

Get the dashboard definition out of Datadog in any form the assistant can read:

The Datadog UI: open the dashboard, then ⚙ Settings → Export dashboard JSON.
The Datadog API: GET /api/v1/dashboard (all dashboards) or GET /api/v1/dashboard/{dashboard_id}.
A Terraform export of your datadog_dashboard resources.

Step 2 — Generate the KubeSense dashboard

Ask the assistant to convert the dashboard, pasting the exported JSON:

Convert this Datadog dashboard into a KubeSense dashboard:
{ …paste the exported Datadog dashboard JSON here… }

The assistant will:

Discover the real names — it queries your telemetry through the MCP server to find the KubeSense metric and label names that correspond to each Datadog query (Datadog kubernetes.* / kubernetes_state.* metrics map to KubeSense kube_*, k8s_pod_*, and container_* metrics).
Map each widget — widget type to panel type, query aggregators and by groupings to PromQL functions, tag filters to label filters, and Datadog template variables to dashboard variables.
Emit importable JSON — a single KubeSense dashboard object with name, description, and a preset that lays the panels out on the grid.

Step 3 — Import and review

In KubeSense, open the Dashboards page and click Import. Paste the generated JSON, or upload a .json file. The dashboard opens with its panels laid out — open each panel to confirm it returns data, and adjust filters, variables, or the Y-axis format if needed.

Nothing is written to your account until you save the imported dashboard.

What converts

Datadog widget	KubeSense equivalent
`timeseries`	Time Series panel
`query_value`	Stat panel
`toplist`	Top List panel
`query_table`	Table panel
`change` (week-over-week)	Stat panel with percentage change — KubeSense has no `calendar_shift`
`note` / `free_text`	No panel equivalent — folded into the dashboard description
`group` (section container)	A collapsible sub-grid (row), or its child widgets flattened into the grid

Always review before saving

Whether you convert one dashboard or many, review it after import. Open each panel to confirm it renders against your live data, and adjust filters, variables, or formatting as needed — so you keep only the panels you have confirmed.