Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
Apps provide visibility at three levels:
| Level | Description |
|---|---|
| App telemetry | Runtime metrics from your running app, which you forward to Application Insights or any monitoring tool you choose. |
| Platform monitoring | Build and deploy health from the CLI, plus tenant-wide usage analytics and operational health in the Microsoft 365 admin center. |
| CLI telemetry | Anonymized usage and diagnostic data from the ms CLI itself, which you can turn off. |
App telemetry
The Microsoft Copilot Managed Runtime SDK reports runtime metrics, such as session load performance and network request summaries, to a logger you provide. You decide where those metrics go. Azure Monitor Application Insights is a common choice, but the same pattern works for any monitoring tool. The platform delivers the metric payloads; your logger implementation decides how to handle them.
Note
The app captures telemetry only after it loads successfully. Startup failures, such as an app blocked from loading, don't appear here. They surface in platform monitoring.
Set up Application Insights
Create an Application Insights resource in the Azure portal and copy its connection string.
Install the Application Insights web SDK in your project.
npm install @microsoft/applicationinsights-webInitialize Application Insights in your app.
import { ApplicationInsights } from '@microsoft/applicationinsights-web'; const appInsights = new ApplicationInsights({ config: { connectionString: 'InstrumentationKey=<YOUR_KEY>;IngestionEndpoint=<YOUR_ENDPOINT>', }, }); appInsights.loadAppInsights();Register a logger with the Copilot Managed Runtime SDK so the platform forwards metrics to Application Insights. Register it once.
import { setConfig } from '@microsoft/cmr-apps'; import type { Metric } from '@microsoft/cmr-apps'; setConfig({ logger: { logMetric: (value: Metric) => { appInsights.trackEvent({ name: value.type }, value.data); }, }, });If Content Security Policy (CSP) is enforced in your environment, the default policy blocks telemetry from reaching Application Insights, because
connect-srcallows only your app's own origin. Have an administrator add the Application Insights ingestion origins toconnect-src. See Configure Content Security Policy to identify the blocked origins and update the directive.
View the data
In your Application Insights resource, open Monitoring > Logs and query the customEvents table to see the metrics your app sends. The SDK reports built-in metrics for session load and network requests, and you can log your own events by calling the Application Insights APIs directly at the points you care about.
Important
Avoid logging sensitive or unnecessary data, and always follow your organization's compliance guidelines.
Platform monitoring
Beyond per-app telemetry, administrators can see every app in the tenant in the Microsoft 365 admin center. They get operational health metrics with alerting to detect and resolve problems before they affect users.
These experiences require no setup from the developer. Apps automatically appear in the centralized inventory and monitoring.
Learn how to add alerts for your apps
CLI telemetry
The ms CLI sends anonymized telemetry by default to help improve the product. It captures usage and diagnostic signals, such as which commands run, whether they succeed, and performance and error data. It doesn't collect your source code or your app's data.
Use the ms telemetry commands to control collection.
Check the current setting with ms telemetry status:
ms telemetry status
Turn telemetry off with ms telemetry disable:
ms telemetry disable
Turn it back on with ms telemetry enable:
ms telemetry enable
Both enable and disable accept two flags that select what they apply to:
--remote: Controls sending telemetry to the service. This flag is the default when no flag is given.--console: Controls printing telemetry events to your local console, which is useful for seeing exactly what the CLI would emit.
Pass both flags to toggle both targets at once.
Build and deploy telemetry
Every build the platform runs produces a record you can inspect from the CLI. Use ms app build-status to check the status and outcome of a build:
ms app build-status --app <name> --commit <sha>
If a build failed, the output includes the reason, so you can fix the code and push again. Build status describes that build, not what's currently live, and it doesn't perform a rollback. A successful build isn't proof that its commit is deployed.
Use ms app info to inspect current server-side app state, including the deployed commit abbreviated to seven characters.
To restore a known-good version, use the deployment command for your app's repository mode:
- Git-backed app: Redeploy a known-good, pushed commit with
ms app deploy --commit <sha>. See Revert to a previous version. - External-artifact app: Redeploy a retained known-good build artifact with
ms app deploy --artifact <path>. These apps don't support--commit. External artifact deployment requires the relevant environment setting.