Security & Backup

Datadog

Bring your Datadog observability into Dex - browse monitors, incidents, hosts, and metrics in plain language.

Dex connects to Datadog through the HTTP API so admins can review observability data conversationally - list and read monitors, browse incidents, inventory hosts, and check the metrics catalog and org metadata for one Datadog organization. This is a read-focused surface: Dex summarizes what Datadog has recorded but does not create, mute, or schedule downtime. Each Dex connection uses one Access Token scoped to a single organization on a single Datadog site (region), so the token must match the org and region it came from.

Start free →See Dex Live

What Dex does with Datadog

Dex handles both admin workflows and employee self-service — all policy-guardrailed and audit-logged.

For admins (Dex Pro)

List Datadog monitors and read a single monitor by id
Browse incidents and read a single incident by id
Inventory hosts across the organization
Check the metrics catalog by name
Read organization metadata (requires the org_read scope; otherwise reported as a scope limit)

Just ask Dex

Your team types a request in plain language. Dex investigates, plans, and executes — with the right guardrails.

Admin prompts

>List the Datadog monitors in this organization
>Show me the details for monitor 123456
>What incidents are currently recorded in Datadog?
>Read incident abc-incident-id and summarize it
>How many hosts is Datadog tracking right now?

Policy actions

Every action Dex can take on Datadog is declared, scoped, and guardrailed. Admins control which apply, who approves them, and whether they're limited to self-service.

Action	What it does	Guardrails
`datadog_list_monitors`	List Datadog monitors via GET /api/v1/monitor	—
`datadog_get_monitor`	Read one Datadog monitor by id	—
`datadog_list_incidents`	List Datadog incidents via GET /api/v2/incidents	—
`datadog_get_incident`	Read one Datadog incident by id	—
`datadog_list_hosts`	List Datadog hosts via GET /api/v1/hosts	—
`datadog_list_metrics`	List Datadog metric names via GET /api/v1/metrics	—
`datadog_get_org`	Read Datadog organization metadata via GET /api/v1/org (403 without org_read - optional surface)	—

How to configure Datadog

Onboarding takes minutes. Dex validates your credentials before saving them.

Setup steps

1
In Datadog, open Organization Settings > Access Tokens (or Personal Settings > My Access Tokens).
2
Create or select a token, then click Copy token - the value starts with "ddpat_" (copy the value, not the token id shown beside it).
3
Grant the token the read scopes Dex needs: monitors_read, hosts_read, incidents_read, and metrics_read (add org_read for org metadata).
4
Note your Datadog site slug (us1, us3, us5, eu, ap1, ap2, or us1-fed) - it must match the region of the org the token came from.
5
In Dex, enable the Datadog integration and paste the site slug and Access Token. Dex validates the token with a live GET /api/v1/monitor call before saving.

Credentials required

site: Datadog site slug (us1, us3, us5, eu, ap1, ap2, us1-fed) or an allowlisted API host - must match the org region; defaults to us1
access_token: Datadog Access Token (value starts "ddpat_"); sent as Authorization: Bearer, never the token id

Requirements

•A Datadog Access Token with the read scopes monitors_read, hosts_read, incidents_read, and metrics_read (org_read is optional for org metadata)
•The site slug must match the region of the organization the token was issued in - a token from another region or org will fail
•This is a per-tenant app: one Access Token per Datadog organization, with no MSP shared-master or child-org model
•Read-only for now - Dex browses observability data but does not create, mute, or schedule downtime