DataDog · rtrieu · Nov 10, 2025 · Nov 10, 2025 · Nov 10, 2025 · Nov 11, 2025
@@ -0,0 +1,182 @@
+---
+title: Agentic Onboarding Setup
+description: Instrument your frontend applications with one prompt using LLM coding agents like Cursor or Claude.
+further_reading:
+
+---
+
+{{< callout btn_hidden="true" header="Join the Preview!">}}
+Agentic Onboarding is in Preview.
+{{< /callout >}}
+
+{{< site-region region="gov" >}}
+<div class="alert alert-danger">Agentic Onboarding is not available in the selected site ({{< region-param key="dd_site_name" >}}) at this time.</div>
+{{< /site-region >}}
+
+## Overview
+
+Datadog's Agentic Onboarding allows you to instrument your frontend applications with one prompt using LLM coding agents like [Cursor][1] or [Claude][2].
+
+Instead of navigating multiple setup steps or searching through documentation, you can instrument your frontend applications for [Error Tracking][3], [Real User Monitoring (RUM)][4], and [Product Analytics][5] in one command.
+
+With Agentic Onboarding, your coding assistant automatically detects your project's frameworks, adds the necessary configuration, and creates the required tokens and apps, all without leaving your IDE.
+
+## Prerequisites
+### Supported frameworks
+Agentic Onboarding is available for the following frameworks: React, Next.js, Vue, Svelte, Angular, Vanilla JS, iOS, Android
+
+## Setup
+
+### Install the Datadog Onboarding MCP server
+
+To install the Datadog Onboarding Model Context Protocol (MCP) server, follow the steps below.
+
+1. Copy the deeplink or command for your AI agent client:
+
+   {{< tabs >}}
+   {{% tab "Cursor" %}}
+
+   Paste the following Cursor deeplink into your browser.
+
+   {{% site-region region="us" %}}
+
+   ```shell
+   cursor://anysphere.cursor-deeplink/mcp/install?name=datadog-onboarding-mcp&config=eyJ1cmwiOiJodHRwczovL21jcC5kYXRhZG9naHEuY29tL2FwaS91bnN0YWJsZS9tY3Atc2VydmVyL21jcD90b29sc2V0cz1vbmJvYXJkaW5nIiwidHlwZSI6Im9hdXRoIn0=
+   ```
+   {{% /site-region %}}
+
+   {{% site-region region="us3" %}}
+   ```shell
+   cursor://anysphere.cursor-deeplink/mcp/install?name=datadog-onboarding-us3&config=eyJ1cmwiOiJodHRwczovL21jcC51czMuZGF0YWRvZ2hxLmNvbS9hcGkvdW5zdGFibGUvbWNwLXNlcnZlci9tY3A/dG9vbHNldHM9b25ib2FyZGluZyIsInR5cGUiOiJvYXV0aCJ9
+   ```
+   {{% /site-region %}}
+
+   {{% site-region region="us5" %}}
+   ```shell
+   cursor://anysphere.cursor-deeplink/mcp/install?name=datadog-onboarding-us5&config=eyJ1cmwiOiJodHRwczovL21jcC51czUuZGF0YWRvZ2hxLmNvbS9hcGkvdW5zdGFibGUvbWNwLXNlcnZlci9tY3A/dG9vbHNldHM9b25ib2FyZGluZyIsInR5cGUiOiJvYXV0aCJ9
+   ```
+   {{% /site-region %}}
+
+   {{% site-region region="eu" %}}
+   ```shell
+   cursor://anysphere.cursor-deeplink/mcp/install?name=datadog-onboarding-mcp&config=eyJ1cmwiOiJodHRwczovL21jcC5kYXRhZG9naHEuZXUvYXBpL3Vuc3RhYmxlL21jcC1zZXJ2ZXIvbWNwP3Rvb2xzZXRzPW9uYm9hcmRpbmciLCJ0eXBlIjoib2F1dGgifQ==
+   ```
+   {{% /site-region %}}
+
+   {{% site-region region="ap1" %}}
+   ```shell
+   cursor://anysphere.cursor-deeplink/mcp/install?name=datadog-onboarding-ap1&config=eyJ1cmwiOiJodHRwczovL21jcC5hcDEuZGF0YWRvZ2hxLmNvbS9hcGkvdW5zdGFibGUvbWNwLXNlcnZlci9tY3A/dG9vbHNldHM9b25ib2FyZGluZyIsInR5cGUiOiJvYXV0aCJ9
+   ```
+   {{% /site-region %}}
+
+   {{% site-region region="ap2" %}}
+   ```shell
+   cursor://anysphere.cursor-deeplink/mcp/install?name=datadog-onboarding-ap2&config=eyJ1cmwiOiJodHRwczovL21jcC5hcDIuZGF0YWRvZ2hxLmNvbS9hcGkvdW5zdGFibGUvbWNwLXNlcnZlci9tY3A/dG9vbHNldHM9b25ib2FyZGluZyIsInR5cGUiOiJvYXV0aCJ9
+   ```
+   {{% /site-region %}}
+
+   {{< site-region region="gov" >}}
+   <div class="alert alert-danger">Agentic Onboarding is not available in the selected site ({{< region-param key="dd_site_name" >}}) at this time.</div>
+   {{< /site-region >}}
+
+   {{% /tab %}}
+
+   {{% tab "Claude Code" %}}
+
+   Copy and execute the Claude Code command into your terminal:
+
+   {{% site-region region="us" %}}
+   ```shell
+   claude mcp add --transport http datadog-onboarding-us1 "https://mcp.datadoghq.com/api/unstable/mcp-server/mcp?toolsets=onboarding"
+   ```
+   {{% /site-region %}}
+
+   {{% site-region region="us3" %}}
+   ```shell
+   claude mcp add --transport http datadog-onboarding-us3 "https://mcp.us3.datadoghq.com/api/unstable/mcp-server/mcp?toolsets=onboarding"
+   ```
+   {{% /site-region %}}
+
+   {{% site-region region="us5" %}}
+   ```shell
+   claude mcp add --transport http datadog-onboarding-us5 "https://mcp.us5.datadoghq.com/api/unstable/mcp-server/mcp?toolsets=onboarding"
+   ```
+   {{% /site-region %}}
+
+   {{% site-region region="eu" %}}
+   ```shell
+   claude mcp add --transport http datadog-onboarding-eu1 "https://mcp.datadoghq.eu/api/unstable/mcp-server/mcp?toolsets=onboarding
+   ```
+   {{% /site-region %}}
+
+   {{% site-region region="ap1" %}}
+   ```shell
+   claude mcp add --transport http datadog-onboarding-ap1 "https://mcp.ap1.datadoghq.com/api/unstable/mcp-server/mcp?toolsets=onboarding"
+   ```
+   {{% /site-region %}}
+
+   {{% site-region region="ap2" %}}
+   ```shell
+   claude mcp add --transport http datadog-onboarding-ap2 "https://mcp.ap2.datadoghq.com/api/unstable/mcp-server/mcp?toolsets=onboarding"
+   ```
+   {{% /site-region %}}
+
+   {{< site-region region="gov" >}}
+   <div class="alert alert-danger">Agentic Onboarding is not available in the selected site ({{< region-param key="dd_site_name" >}}) at this time.</div>
+   {{< /site-region >}}
+
+   {{% /tab %}}
+   {{< /tabs >}}
+
+2. In your AI agent client, click **Install** for the `datadog-onboarding-mcp` server.
+3. If you see a **Needs login** link under the installed MCP server, click the link to complete the Oauth process.
+4. When prompted to open an external website, click **Open**.
+5. After you've granted access to your Datadog account, you are redirected to Cursor. Click **Open** to complete the authentication process.
+5. Confirm you see MCP tools listed for the `datadog-onboarding-mcp` server.
+
+### Set up your project
+
+Your AI coding agent can automatically configure Datadog for your project. When you provide a setup prompt, your coding agent does the following:
+
+- Analyze your project and identify if the MCP server offers a tool that can be used to set it up with Datadog
+- Call the tool (asking for your permission before doing so) with inferred parameters from your project (for example: your project's framework, language, and bundler)
+- Follow the instructions the MCP tool provides as context to your coding agent, making code changes on your behalf (don't worry - Datadog does not commit them)
+- Provide testing steps to confirm that your application is correctly configured to send telemetry to Datadog
+
+1. To get started, copy and paste the following prompt based on the product you want to use into your coding agent (such as Cursor or Claude Code):
+
+   {{< tabs >}}
+   {{% tab "Error Tracking" %}}
+   ```console
+   Add Datadog Error Tracking to my project
+   ```
+   {{% /tab %}}
+
+   {{% tab "Real User Monitoring" %}}
+   ```console
+   Add Datadog Real User Monitoring to my project
+   ```
+   {{% /tab %}}
+
+   {{% tab "Product Analytics" %}}
+   ```console
+   Add Datadog Product Analytics to my project
+   ```
+   {{% /tab %}}
+   {{< /tabs >}}
+
+2. After pasting the prompt, review and accept each action your AI agent proposes to move through the setup process.
+
+### Deploy your app to production
+
+Depending on how your application is deployed, you need to commit the changes and set or upload provided environment variables to your production environment.
+
+[1]: https://cursor.com/
+[2]: https://claude.ai/
+[3]: /error_tracking/frontend_monitoring
+[4]: /real_user_monitoring/
+[5]: /product_analytics/
+[6]: https://platform.openai.com/docs/guides/text
+[7]: https://github.com/langchain-ai/langgraph
+[8]: https://github.com/vercel/ai-chatbot
+
@@ -23,6 +23,19 @@ Additional features are available depending on the source of the error. See [sup
 - Take a tour of key Error Tracking features in the [Error Tracking Explorer][5] documentation.
 - Use the product-specific links in the next section to set up Error Tracking for a particular error source.
 
+## Setup
+{{< whatsnext desc="To get started with Datadog Error Tracking, see the corresponding documentation:" >}}
+    {{< nextlink href="error_tracking/frontend/browser" >}}Browser{{< /nextlink >}}
+    {{< nextlink href="error_tracking/frontend/browser" >}}Browser{{< /nextlink >}}
+    {{< nextlink href="error_tracking/frontend/mobile/android" >}}Android{{< /nextlink >}}
+    {{< nextlink href="error_tracking/frontend/mobile/ios" >}}iOS{{< /nextlink >}}
+    {{< nextlink href="error_tracking/frontend/mobile/expo" >}}Expo{{< /nextlink >}}
+    {{< nextlink href="error_tracking/frontend/mobile/reactnative" >}}React Native{{< /nextlink >}}
+    {{< nextlink href="error_tracking/frontend/mobile/flutter" >}}Flutter{{< /nextlink >}}
+    {{< nextlink href="error_tracking/frontend/mobile/kotlin_multiplatform" >}}Kotlin Multiplatform{{< /nextlink >}}
+    {{< nextlink href="error_tracking/frontend/logs" >}}Logs{{< /nextlink >}}
+{{< /whatsnext >}}
+
 ## Supported error sources
 
 Error Tracking captures and processes errors across your web, mobile, and backend applications. You can instrument your applications and services using the [Browser SDK][6], [Mobile SDK][7], or ingest errors from your Logs, Traces, and Real User Monitoring events. 

@@ -0,0 +1,6 @@
+---
+title: Agentic Onboarding for RUM
+description: Instrument your frontend application with one prompt using LLM coding agents like Cursor or Claude.
+---
+
+{{< include-markdown "agentic_onboarding/setup" >}}
@@ -27,7 +27,14 @@ The Datadog Android SDK supports Android 6.0+ (API level 23) and Android TV.
 
 ## Setup
 
-To start sending RUM data from your Android or Android TV application to Datadog:
+**Choose your setup method:**
+
+- **[Agentic Onboarding (in Preview)][18]**: Use AI coding agents (Cursor, Claude Code) to automatically instrument your Android application with one prompt. The agent detects your project structure and configures the RUM SDK for you.
+- **Manual setup** (below): Follow the step-by-step instructions to manually add and configure the RUM SDK in your Android application.
+
+### Manual setup
+
+To start sending RUM data from your Android or Android TV application to Datadog, follow the steps below.
 
 ### Step 1 - Declare the Android SDK as a dependency
 
@@ -637,3 +644,4 @@ val inputStream = context.getRawResAsRumResource(id)
 [15]: https://square.github.io/okhttp/features/interceptors/#network-interceptors
 [16]: /real_user_monitoring/application_monitoring/android/advanced_configuration/#automatically-track-network-requests
 [17]: https://square.github.io/okhttp/features/interceptors/
+[18]: /real_user_monitoring/application_monitoring/agentic_onboarding/?tab=realusermonitoring
@@ -12,22 +12,25 @@ further_reading:
 ## Setup
 
 {{< whatsnext desc="Choose the instrumentation type for the Browser SDK:" >}}
-  {{< nextlink href="real_user_monitoring/application_monitoring/browser/setup/client">}}<u>Client-Side</u>: Instrument each of your browser-based web applications, deploy the application, then configure the initialization parameters you want to track, and use advanced configuration to further manage data and context that RUM collects.{{< /nextlink >}}
   {{< nextlink href="/real_user_monitoring/application_monitoring/browser/setup/server">}}<u>Auto-Instrumentation</u>: Inject a RUM SDK JavaScript scriptlet into the HTML responses of your web applications being served through a web server or proxy.{{< /nextlink >}}
+  {{< nextlink href="/real_user_monitoring/application_monitoring/agentic_onboarding">}}<u>Agentic Onboarding</u>: Instrument your frontend applications with one prompt using LLM coding agents like Cursor or Claude.{{< /nextlink >}}
+  {{< nextlink href="real_user_monitoring/application_monitoring/browser/setup/client">}}<u>Client-Side</u>: Instrument each of your browser-based web applications, deploy the application, then configure the initialization parameters you want to track, and use advanced configuration to further manage data and context that RUM collects.{{< /nextlink >}}
 {{< /whatsnext >}}
 
 ## How to choose the instrumentation type
 
-| | Auto-instrumentation (Preview) | Client-side (Manual) |
-|----------------------|--------------------------------|----------------------|
-| **SDK setup mechanism** | [Automatically][1] add RUM JS to your web app HTML. Once RUM Auto-instrumentation is set-up, manage configurations from the UI. | [Manually][2] add the RUM SDK to your application code and manage configurations in code. |
-| **Code changes required** | No | Yes |
-| **Setup complexity** | Low | Medium |
-| **User groups** | **SRE and engineering teams** without access to frontend code, or **teams who need to manage** all observability needs centrally, may find this useful for: <br>  - Unlocking performance data across all applications upon setting up RUM <br>  - Holistically monitoring application performance across the organization | **Frontend engineering, mobile engineering, or product teams** with access to frontend code may find this method useful for: <br>  - Daily engineering needs (for example: live support, troubleshooting, and health checks for downstream services) <br>  -Product needs (for example: user flow analysis, user segmentation, and feature flag tracking) <br>  - Capturing observability from in-house code or complex functions that aren't captured by automatic instrumentation |
+| | Auto-instrumentation (Preview) | Agentic Onboarding (Preview) | Client-side (Manual) |
+|----------------------|--------------------------------|------------------------------|----------------------|
+| **SDK setup mechanism** | [Automatically][1] add RUM JS to your web app HTML. Once RUM Auto-instrumentation is set-up, manage configurations from the UI. | [AI-guided setup][3] that automatically detects your project's framework and adds the RUM SDK with one prompt using coding agents. | [Manually][2] add the RUM SDK to your application code and manage configurations in code. |
+| **Code changes required** | No | Yes (automated by AI agent) | Yes |
+| **Setup complexity** | Low | Low | Medium |
+| **Supported platforms** | Apache, IBM HTTP Server, Java Servlet, Nginx, Windows IIS | Next.js, React, Svelte, Vue, Vanilla JavaScript | All browser-based applications |
+| **User groups** | **SRE and engineering teams** without access to frontend code, or **teams who need to manage** all observability needs centrally, may find this useful for: <br>  - Unlocking performance data across all applications upon setting up RUM <br>  - Holistically monitoring application performance across the organization | **Teams using AI coding agents** (Cursor, Claude Code) may find this useful for: <br>  - Accelerating RUM setup with AI-guided instrumentation <br>  - Automating framework detection and SDK configuration <br>  - Reducing time-to-observability for new projects | **Frontend engineering, mobile engineering, or product teams** with access to frontend code may find this method useful for: <br>  - Daily engineering needs (for example: live support, troubleshooting, and health checks for downstream services) <br>  -Product needs (for example: user flow analysis, user segmentation, and feature flag tracking) <br>  - Capturing observability from in-house code or complex functions that aren't captured by automatic instrumentation |
 
 ## Further reading
 
 {{< partial name="whats-next/whats-next.html" >}}
 
 [1]: /real_user_monitoring/application_monitoring/browser/setup/server
 [2]: /real_user_monitoring/application_monitoring/browser/setup/client
+[3]: /real_user_monitoring/application_monitoring/agentic_onboarding/?tab=realusermonitoring
@@ -47,7 +47,14 @@ Before you begin, ensure you have:
 
 ## Setup
 
-To start sending RUM data from your iOS or tvOS application to Datadog:
+**Choose your setup method:**
+
+- **[Agentic Onboarding (in Preview)][14]**: Use AI coding agents (Cursor, Claude Code) to automatically instrument your iOS application with one prompt. The agent detects your project structure and configures the RUM SDK for you.
+- **Manual setup** (below): Follow the step-by-step instructions to manually add and configure the RUM SDK in your iOS application.
+
+### Manual setup
+
+To start sending RUM data from your iOS or tvOS application to Datadog, follow the steps below.
 
 ### Step 1 - Add the iOS SDK as a dependency
 
@@ -553,3 +560,4 @@ See [Supported versions][9] for a list of operating system versions and platform
 [11]: /real_user_monitoring/ios/web_view_tracking/
 [12]: /real_user_monitoring/ios/data_collected/
 [13]: https://app.datadoghq.com/rum/application/
+[14]: /real_user_monitoring/application_monitoring/agentic_onboarding/?tab=realusermonitoring
@@ -41,6 +41,15 @@ The React Native SDK supports the following services:
 
 ## Setup
 
+**Choose your setup method:**
+
+- **[Agentic Onboarding (in Preview)][15]**: Use AI coding agents (Cursor, Claude Code) to automatically instrument your React Native application with one prompt. The agent detects your project structure and configures the RUM SDK for you.
+- **Manual setup** (below): Follow the step-by-step instructions to manually add and configure the RUM SDK in your React Native application.
+
+### Manual setup
+
+To start sending RUM data from your React Native application to Datadog, follow the steps below.
+
 To install with NPM, run:
 
 ```sh
@@ -478,3 +487,4 @@ end
 [12]: https://support.apple.com/guide/security/security-of-runtime-process-sec15bfe098e/web
 [13]: https://stackoverflow.com/questions/37388126/use-frameworks-for-only-some-pods-or-swift-pods/60914505#60914505
 [14]: https://reactnative.dev/architecture/landing-page
+[15]: /real_user_monitoring/application_monitoring/ios/agentic_onboarding/?tab=realusermonitoring