Manual Setup

Prerequisites

You need:

• A Sentry account and project
• Your application up and running
• SvelteKit version 2.0.0+
• Vite version 4.2+

Step 1: Install

Choose the features you want to configure, and this guide will show you how:

Install the Sentry SDK

Run the command for your preferred package manager to add the Sentry SDK to your application:

npm install @sentry/sveltekit --save

If you're updating your Sentry SDK to the latest version, check out our migration guide to learn more about breaking changes.

Step 2: Configure

You need to initialize and configure the Sentry SDK in three places: the client side, the server side, and your Vite config.

Configure Client-Side Sentry

Create a client hooks file src/hooks.client.(js|ts) in the src folder of your project if you don't have one already. In this file, import and initialize the Sentry SDK and add the handleErrorWithSentry function to the handleError hook.

import * as Sentry from "@sentry/sveltekit";

Sentry.init({
  dsn: "https://examplePublicKey@o0.ingest.sentry.io/0",

  // Adds request headers and IP for users, for more info visit:
  // https://docs.sentry.io/platforms/javascript/guides/sveltekit/configuration/options/#sendDefaultPii
  sendDefaultPii: true,
  //  performance

  // Set tracesSampleRate to 1.0 to capture 100%
  // of transactions for tracing.
  // We recommend adjusting this value in production
  // Learn more at
  // https://docs.sentry.io/platforms/javascript/configuration/options/#traces-sample-rate
  tracesSampleRate: 1.0,
  //  performance
  //  session-replay
  integrations: [Sentry.replayIntegration()],

  // Capture Replay for 10% of all sessions,
  // plus for 100% of sessions with an error
  // Learn more at
  // https://docs.sentry.io/platforms/javascript/session-replay/configuration/#general-integration-configuration
  replaysSessionSampleRate: 0.1,
  replaysOnErrorSampleRate: 1.0,
  //  session-replay
});

const myErrorHandler = ({ error, event }) => {
  console.error("An error occurred on the client side:", error, event);
};

export const handleError = Sentry.handleErrorWithSentry(myErrorHandler);

// or alternatively, if you don't have a custom error handler:
// export const handleError = handleErrorWithSentry();

Configure Server-Side Sentry

Create a server hooks file src/hooks.server.(js|ts) in the src folder of your project if you don't have one already. In this file, import and initialize the Sentry SDK and add the handleErrorWithSentry function to the handleError hook and the Sentry request handler to the handle hook.

import * as Sentry from "@sentry/sveltekit";

Sentry.init({
  dsn: "https://examplePublicKey@o0.ingest.sentry.io/0",

  // Adds request headers and IP for users, for more info visit:
  // https://docs.sentry.io/platforms/javascript/guides/sveltekit/configuration/options/#sendDefaultPii
  sendDefaultPii: true,
  //  performance

  // Set tracesSampleRate to 1.0 to capture 100%
  // of transactions for tracing.
  // We recommend adjusting this value in production
  // Learn more at
  // https://docs.sentry.io/platforms/javascript/configuration/options/#traces-sample-rate
  tracesSampleRate: 1.0,
  //  performance
});

const myErrorHandler = ({ error, event }) => {
  console.error("An error occurred on the server side:", error, event);
};

export const handleError = Sentry.handleErrorWithSentry(myErrorHandler);
// or alternatively, if you don't have a custom error handler:
// export const handleError = handleErrorWithSentry();

export const handle = Sentry.sentryHandle();
// Or use `sequence` if you're using your own handler(s):
// export const handle = sequence(Sentry.sentryHandle(), yourHandler());

Configure Vite

Add the sentrySvelteKit plugin before sveltekit in your vite.config.(js|ts) file to automatically upload source maps to Sentry and instrument load functions for tracing if it's configured.

import { sveltekit } from "@sveltejs/kit/vite";
import { sentrySvelteKit } from "@sentry/sveltekit";
import { defineConfig } from "vite";

export default defineConfig({
  plugins: [sentrySvelteKit(), sveltekit()],
  // ... rest of your Vite config
});

Source Maps Upload

By default, sentrySvelteKit() will add an instance of the Sentry Vite Plugin, to upload source maps for both server and client builds. This means that when you run a production build (npm run build), source maps will be generated and uploaded to Sentry, so that you get readable stack traces in your Sentry issues.

However, you still need to specify your Sentry auth token as well as your org and project slugs. There are two ways to set them.

You can set them as environment variables, for example in a .env file:

• SENTRY_ORG your Sentry org slug
• SENTRY_PROJECT your Sentry project slug
• SENTRY_AUTH_TOKEN your Sentry auth token (can be obtained from the Organization Token Settings)
• SENTRY_URL your Sentry instance URL. This is only required if you use your own Sentry instance (as opposed to https://sentry.io).

Or, you can set them by passing a sourceMapsUploadOptions object to sentrySvelteKit, as seen in the example below. All available options are documented at the Sentry Vite plugin repo.

SENTRY_AUTH_TOKEN=sntrys_YOUR_TOKEN_HERE

Using the sourceMapsUploadOptions object is useful if the default source maps upload doesn't work out of the box, for instance, if you have a customized build setup or if you're using the SDK with a SvelteKit adapter other than the supported adapters.

Overriding SvelteKit Adapter detection

By default, sentrySvelteKit will try to detect your SvelteKit adapter to configure source maps upload correctly. If you're not using one of the supported adapters or the wrong one is detected, you can override the adapter detection by passing the adapter option to sentrySvelteKit:

import { sveltekit } from "@sveltejs/kit/vite";
import { sentrySvelteKit } from "@sentry/sveltekit";

export default {
  plugins: [
    sentrySvelteKit({
      adapter: "vercel",
    }),
    sveltekit(),
  ],
  // ... rest of your Vite config
};

Disable Source Maps Upload

You can disable automatic source maps upload in your Vite config:

import { sveltekit } from "@sveltejs/kit/vite";
import { sentrySvelteKit } from "@sentry/sveltekit";

export default {
  plugins: [
    sentrySvelteKit({
      autoUploadSourceMaps: false,
    }),
    sveltekit(),
  ],
  // ... rest of your Vite config
};

If you disable automatic source maps upload, you must explicitly set a release value in your Sentry.init() configs. You can also skip the sentry-cli configuration step below.

Step 3: Optional Configuration

The points explain additional optional configuration or more in-depth customization of your Sentry SvelteKit SDK setup.

Auto-Instrumentation

The SDK primarily uses SvelteKit's hooks to collect error and performance data. However, SvelteKit doesn't yet offer a hook for universal or server-only load function calls. Therefore, the SDK uses a Vite plugin to auto-instrument load functions so that you don't have to add a Sentry wrapper to each function manually.

Auto-instrumentation is enabled by default when you add the sentrySvelteKit() function call to your vite.config.(js|ts). However, you can customize the behavior, or disable it entirely. If you disable it you can still manually wrap specific load functions with the withSentry function.

Customize Auto-instrumentation

By passing the autoInstrument option to sentrySvelteKit you can disable auto-instrumentation entirely, or customize which load functions should be instrumented:

import { sveltekit } from "@sveltejs/kit/vite";
import { sentrySvelteKit } from "@sentry/sveltekit";

export default {
  plugins: [
    sentrySvelteKit({
      autoInstrument: {
        load: true,
        serverLoad: false,
      },
    }),
    sveltekit(),
  ],
  // ... rest of your Vite config
};

Disable Auto-instrumentation

If you set the autoInstrument option to false, the SDK won't auto-instrument any load functions. You can still manually instrument specific load functions.

import { sveltekit } from '@sveltejs/kit/vite';
import { sentrySvelteKit } from '@sentry/sveltekit';

export default {
  plugins: [
    sentrySvelteKit({
      autoInstrument: false;
    }),
    sveltekit(),
  ],
  // ... rest of your Vite config
};

Configure CSP for Client-side fetch Instrumentation

If you're using a SvelteKit version older than sveltejs/kit@2.16.0, The Sentry SDK injects an inline <script> tag into the HTML response of the server. This script proxies all client-side fetch calls, so that fetch calls inside your load functions are captured by the SDK. However, if you configured CSP rules to block inline fetch scripts by default, this script will be blocked by the browser.

You have multiple options to solve this CSP issue:

Upgrade SvelteKit

The easiest option is to update @sveltejs/kit to at least version 2.16.0 (or newer). The SDK will not inject the fetch proxy script as it's no longer necessary.

Add Script Hash to CSP policy

To enable the script, add an exception for the sentryHandle script by adding the hash of the script to your CSP script source rules.

If your CSP is defined in svelte.config.js, you can add the hash to the csp.directives.script-src array:

const config = {
  kit: {
    csp: {
      directives: {
        "script-src": [
          "sha256-y2WkUILyE4eycy7x+pC0z99aZjTZlWfVwgUAfNc1sY8=",
        ], // + rest of your values
      },
    },
  },
};

For external CSP configurations, add the following hash to your script-src directive:

'sha256-y2WkUILyE4eycy7x+pC0z99aZjTZlWfVwgUAfNc1sY8='

We will not make changes to this script any time soon, so the hash will stay consistent.

Disable fetch Proxy Script

If you don't want to inject the script responsible for instrumenting client-side fetch calls, you can disable injection by passing injectFetchProxyScript: false to sentryHandle:

export const handle = sentryHandle({ injectFetchProxyScript: false });

Note that if you disable the fetch proxy script, the SDK will not be able to capture spans for fetch calls made by SvelteKit when your route has a server load function. This also means that potential spans created on the server for these fetch calls will not be connected to the client-side trace.

Manual Instrumentation

Instead or in addition to Auto Instrumentation, you can manually instrument certain SvelteKit-specific features with the SDK:

Instrument load Functions

SvelteKit's universal and server load functions are instrumented automatically by default. If you don't want to use load auto-instrumentation, you can disable it, and manually instrument specific load functions with the SDK's load function wrappers.

Universal load Functions

Use the wrapLoadWithSentry function to wrap universal load functions declared in +page.(js|ts) or +layout.(js|ts)

import { wrapLoadWithSentry } from "@sentry/sveltekit";

export const load = wrapLoadWithSentry(({ fetch }) => {
  const res = await fetch("/api/data");
  const data = await res.json();
  return { data };
});

Server-only load Functions

Or use the wrapServerLoadWithSentry function to wrap server-only load functions declared in +page.server.(js|ts) or +layout.server.(js|ts)

import { wrapServerLoadWithSentry } from "@sentry/sveltekit";

export const load = wrapServerLoadWithSentry(({ fetch }) => {
  const res = await fetch("/api/data");
  const data = await res.json();
  return { data };
});

Instrument Server Routes

You can also manually instrument server (API) routes with the SDK. This is useful if you have custom server routes that you want to trace or if you want to capture error() calls within your server routes:

import { wrapServerRouteWithSentry } from "@sentry/sveltekit";

export const GET = wrapServerRouteWithSentry(async () => {
  // your endpoint logic
  return new Response("Hello World");
});

Step 4: Cloudflare Pages Configuration (Optional)

If you're deploying your application to Cloudflare Pages, you need to adjust your server-side setup. Follow this guide to configure Sentry for Cloudflare.

Step 5: Avoid Ad Blockers With Tunneling (Optional)

You can prevent ad blockers from blocking Sentry events using tunneling. Use the tunnel option to add an API endpoint in your application that forwards Sentry events to Sentry servers.

Update Sentry.init in your hooks.client.(js|ts) file with the following option:

Sentry.init({
  dsn: "https://examplePublicKey@o0.ingest.sentry.io/0",,
  tunnel: "/tunnel",
});

This will send all events to the tunnel endpoint. However, the events need to be parsed and redirected to Sentry, so you need to do additional configuration on the server. You can find a detailed explanation on how to do this on our Troubleshooting page.

Step 6: Verify Your Setup

Let's test your setup and confirm that Sentry is working correctly and sending data to your Sentry project.

Issues

To verify that Sentry captures errors and creates issues in your Sentry project, create a test page, for example, at src/routes/sentry-example/+page.svelte with a button that throws an error when clicked:

<script>
  function throwTestError() {
    throw new Error("Sentry Example Frontend Error");
  }
</script>

<button type="button" onclick="{throwTestError}">Throw error</button>

export const GET = async () => {
  throw new Error("Sentry Example API Route Error");
};

<script>
  import * as Sentry from "@sentry/sveltekit";

  function throwTestError() {
    Sentry.startSpan(
      {
        name: "Example Frontend Span",
        op: "test",
      },
      async () => {
        const res = await fetch("/sentry-example");
        if (!res.ok) {
          throw new Error("Sentry Example Frontend Error");
        }
      },
    );
  }
</script>

<button type="button" onclick="{throwTestError}">
  Throw error with trace
</button>

View Captured Data in Sentry

Now, head over to your project on Sentry.io to view the collected data (it takes a couple of moments for the data to appear).

Next Steps

At this point, you should have integrated Sentry into your SvelteKit application and should already be sending data to your Sentry project.

Now's a good time to customize your setup and look into more advanced topics. Our next recommended steps for you are:

• Learn how to manually capture errors
• Continue to customize your configuration
• Get familiar with Sentry's product features like tracing, insights, and alerts