Full read-write Claude connector for Datadog
Let Claude manage monitors, mute alerts, and update incidents in Datadog
Stop tab-switching between Claude and Datadog to triage. Claude queries your metrics, logs, and monitor status, then prepares the monitor edits, alert mutes, and incident updates right where your telemetry lives, applying them once you approve.
Datadog is a monitoring and observability platform for metrics, logs, traces, and alerts. Teams want Claude to do more than read dashboards: to mute alerts, manage monitors, and post to incidents directly from a conversation.
What is a Datadog MCP server?
A Datadog MCP server is a small hosted service that exposes Datadog's data and actions to Claude over the Model Context Protocol, so Claude can manage issues, code, and incidents inside your own account. We build it as a custom connector with separate read and write tools, and destructive actions stay gated behind your approval.
What can Datadog's current connector do today?
Today, Datadog offers an official Claude connector in the directory, and it covers a standard tool set by design. That general set will not always match the deeper actions, custom monitor logic, workflow shaping, and approval gates on-call work needs. A custom Datadog connector adds that fuller coverage with scoped, human-in-the-loop controls.
What does full read-write Datadog access unlock?
Full read-write Datadog access lets Claude create and edit monitors, mute or unmute monitors during a deploy, add notes and status updates to an active incident, and adjust dashboards. The triage and on-call follow-up that used to mean tab-switching happen where your telemetry already lives, once you approve.
- Create or edit a monitor with thresholds and notification targets
- Mute an alert or downtime window during a planned deploy
- Post a status update or note to an active incident
- Mute or unmute a monitor during a maintenance window
- Add a widget or update a dashboard from a saved query
- Tag hosts or services to keep monitor scopes accurate
Can Claude create or edit monitors in Datadog?
Yes. Claude can create or edit a monitor in Datadog, setting the thresholds and notification targets you specify and tagging hosts or services to keep scopes accurate. The default read-leaning setup is fine for querying status, so a custom connector adds these write tools, and changes apply once you approve.
Can Claude mute alerts during a deploy in Datadog?
Claude can mute an alert or set a downtime window in Datadog during a planned deploy, then unmute the monitor afterward. Muting, downtimes, and unmuting are separate, gated write actions, so Claude prepares the change and shows you exactly what it will do before it takes effect.
Can Claude post updates to an active incident in Datadog?
Claude can post a status update or note to an active Datadog incident and adjust dashboards from a saved query. The connector inherits exactly the permissions your API and application keys carry and nothing more, so Claude never has more access than the keys you provision, and outbound updates wait for your confirmation.
How do you connect Datadog to Claude?
We build a small MCP server that authenticates to Datadog with scoped API and application keys, requesting only the permissions the work needs. Read tools (query metrics, search logs, list monitors) and write tools (create monitor, mute alert, update incident) are separated, and destructive or outbound actions require your confirmation. Claude never has more access than the keys you provision, and the server can be IP-allowlisted to your environment.
Frequently asked questions
Can Claude create or edit Datadog monitors, not just read them?
Will Claude mute alerts or change incidents without my approval?
Does the connector get more access than my Datadog keys allow?
Using Claude Code?
The same connector works in the terminal: connect Datadog to Claude Code.
Related connectors
Tell us what you run. We'll build the connector.
Pick your tool, choose what Claude should do, and get an instant scope and price. No call required.