Repository navigation
Datadog Shipped - 11/6/26 #6
blondecostanza
announced in
Announcements
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
Kicking off this roundup. 8 ships worth knowing about:
What shipped, why it exists, docs, and more.
1. Datadog auto-connect for Claude is now in beta
What shipped: Your team can now connect to Datadog in Claude without everyone going through their own authentication flow. Admins configure Datadog once for their organization, and eligible teammates connect automatically the first time they use it in Claude, no individual sign-in screens and no separate identity provider required.
Why it exists: The standard way to connect an AI client to a tool is a per-user OAuth flow: every person on a team authorizes separately, every time. That's fine for one person, but it doesn't scale to a whole org, and it leaves no central record of who's connected or how. Auto-connect moves that decision to the admin level, so a team gets set up once instead of one person at a time. It's built on the same underlying technology as Cross-App Access (XAA), but unlike XAA, it doesn't require Okta or any identity provider at all, Claude vouches for each signed-in member directly.
How it works: An admin sets up the connection once from the Claude side; team members with a matching email in both systems get connected automatically the first time they use the Datadog MCP server in Claude. Available for eligible Claude Team and Enterprise plans.
Shipped by: Anthony Dagneaux and Aditya Balasubramanian
Docs: Setup guide: Configure Cross-App Access in Datadog · Anthropic's support article
Start now: Organization Settings in Datadog
Claude.Autoconnect.Walkthrough.mp4
2. Datadog Canvas: production investigations, now multiplayer
What shipped: Datadog Canvas is a collaborative workspace for investigating production issues: start with a question, explore your infrastructure with AI-powered insights, and work alongside teammates and AI agents in real time, all on one shared view that keeps charts, service maps, evidence, and next steps connected together. It's live now in preview.
Why it exists: An investigation usually produces a trail of scattered artifacts: a few dashboards, a Slack thread, some screenshots, a doc someone started and never finished. None of it stays connected, and nobody joining partway through has an easy way to catch up. Canvas keeps the whole investigation, the telemetry, the evidence, the reasoning, and the next steps, in one place that updates live as people and agents work in it together.
Where it stands: In preview, open to design partners and customers who request access.
Shipped by: Reva Ranka
Start now: Request access at canvas.datadoghq.com
Canvas.Walkthrough.mp4
3. Datadog Agent Observability is now live on the Claude Marketplace
What shipped: Datadog Agent Observability is now available directly from the Claude Marketplace, with Datadog featured as an Anthropic launch partner.
Why it exists: Datadog Agent Observability turns real-world usage into a continuous improvement loop for AI agents. Teams building on the Claude Platform can send agent traces to Datadog through OpenTelemetry to automatically detect issues, get suggested fixes, and measure their impact in production, shipping faster while continuously improving reliability, security, user experience, and cost efficiency at scale.
Shipped by: Richard Liu, Mike Cannady, Uday Tennety and Marlisa Gashaj
Docs: Datadog Agent Observability documentation
Blog: Datadog is now available on Anthropic's Claude Marketplace
Start now: Datadog Agent Observability on the Claude Marketplace
4. Three new RUM SDKs: C/C++, Electron, and .NET MAUI
What shipped: Full RUM coverage, wherever your code runs, desktop apps, games, embedded devices, and cross-platform mobile. 3 new SDKs are generally available today:
Why it exists: Each of these platforms represents a category of software RUM previously had partial or no visibility into: native desktop and game clients, Electron apps with real logic running in the Node main process (not just the renderer), and cross-platform .NET mobile apps that previously depended on community bindings or custom wrappers.
Shipped by: Quentin Sellem (C/C++), Maël Lilensten (Electron, .NET MAUI)
Start now: dd-sdk-cpp on GitHub · electron-sdk on GitHub · dd-sdk-maui on GitHub
Docs: C/C++ docs · C/C++ blog · Electron docs · Electron blog · .NET MAUI docs · .NET MAUI blog
5. Pinpoint the root cause of iOS and Android slowdowns with RUM Mobile Profiling
What shipped: RUM Mobile Profiling is now generally available for iOS and Android. It captures code-level CPU profiles on real users' devices and attaches them directly to RUM events, giving mobile teams visibility into exactly which code slowed an app launch or froze the main thread.
Why it exists: Time to Initial Display tells you a launch is slow, it doesn't tell you why. Previously, finding the actual cause meant reviewing individual sessions, searching the codebase, and trying to reproduce the issue locally, difficult when it only happens on specific devices or OS versions. RUM Mobile Profiling captures the code that ran on the user's own device when the app was slow or frozen, so teams no longer depend on test devices to find the cause.
How it works: Profiles attach to app launches (TTID/TTFD), ANRs, app hangs, long tasks, and critical RUM operations. An aggregated profiling page ranks the slowest methods across sessions, with a comparison view for catching regressions between app versions. Investigate from any profiled event, or ask Bits AI for the root cause directly.
Shipped by: Arti Arutiunov
Start now: Aggregated RUM Profiling page
Docs: Android setup · iOS setup · Blog post
6. Monitor any data processing job with OpenLineage
What shipped: Data Observability's Jobs Monitoring now supports OpenLineage, an open standard for data pipeline telemetry. Customers can emit custom OpenLineage run events from any legacy or custom data processing job and get the same alerting, execution traces, and end-to-end lineage that built-in integrations like Spark, Databricks, and Airflow already provide.
Why it exists: Built-in integrations cover the common data stack technologies, but a lineage graph is only as strong as its weakest link, and that's usually a custom script or legacy system nobody got around to instrumenting. Teams would find out their lineage had a gap from the analyst whose dashboard broke, always too late. OpenLineage support means any job that can be instrumented to emit events gets the same visibility as natively-supported ones.
How it works: Send JSON OpenLineage run events over HTTP (START, RUNNING, then COMPLETE or FAIL, sharing one runId) for each job run. Datadog assembles them into run history, status, and lineage, the same way it does for natively-supported technologies.
Shipped by: Ryan Warrier
Start now: See custom jobs in the demo org
Docs: Custom jobs setup · OpenLineage standard
7. Monitor the health of large systems at a glance with the new Heatgrid widget
What shipped: A new Heatgrid widget is available in dashboards and notebooks, letting you spot trends and patterns across many groups at once in a compact, color-coded grid.
Why it exists: Comparing 10 or more services, regions, or hosts over time in a standard timeseries chart gets crowded fast, overlapping lines make it hard to read. The Heatgrid makes that comparison legible, and gives customers a native, configurable option where they previously had to hand-build the same view with the Wildcard widget using raw Vega-Lite JSON.
How it works: Add a Heatgrid to any dashboard or notebook through the native Datadog editor. Supports editable row label widths, brush-to-zoom, and interaction with other widgets on the same dashboard.
Shipped by: Lilia Ennouri
Start now: See a live demo dashboard
Docs: Release note
8. Your dashboards just got more room to breathe: iPhone Duo support for Datadog Mobile
What shipped: The Datadog mobile app now supports iPhone Duo. Unfold the phone and dashboards, monitors, and incident timelines expand to fill the extra space. Fold it back up and pick up right where you left off.
**Why it exists:**Datadog Mobile now takes advantage of the new space in foldable iPhone Duo, specifically for the views people actually check on the go: dashboards, monitors, and incident timelines, same telemetry, more screen to work with.
How it works: Update to the latest version of the Datadog mobile app.
Shipped by: Salim Bouhorma Mouffak
Start now: App Store · Google Play
Docs: Datadog Mobile App documentation
DUO.VIDEO.mp4
Questions or feedback on any of these? Drop a comment below.
All reactions