Migrate from Sentry

Connect a Sentry organization

Use the Sentry App connection when you want to keep Sentry active while Squasher receives issue events and reads selected incident context.

In Squasher, open Integrations > Sentry and select Install Sentry App.
Select the Sentry organization and projects to connect.
Approve read access for organizations, projects, teams, and events.
Use Verify to check the saved installation and webhook configuration.

The Sentry App must be registered and available to the customer organization. Squasher verifies signed Sentry webhooks, de-duplicates deliveries, and marks the connection for repair when an installation is removed. Verify does not create a Sentry issue. Credential-gated live checks use a dedicated Sentry test organization during release verification.

Disconnecting in Squasher removes the stored grant and stops Squasher from processing new Sentry events. Sentry does not provide a public API that lets Squasher uninstall one organization installation. To stop Sentry from sending the installation webhooks, also uninstall the Squasher Sentry App in your Sentry organization settings.

The connected agent operations are typed, read-only Sentry actions. They do not expose a general Sentry HTTP client. Use the DSN migration below when you want to send SDK errors directly to Squasher instead of keeping the Sentry project as the event source.

Squasher is compatible with Sentry SDK error events. You can keep using your existing @sentry/node, sentry-python, sentry-go, sentry-ruby, or any other official Sentry SDK and point its error traffic at Squasher. No new error SDK or capture-call rewrite is required. Update the DSN, disable Sentry transaction sampling if you enabled it, and your errors flow to Squasher with AI triage included on every plan.

How It Works

Sentry SDKs send error data to a configurable endpoint using the Sentry envelope protocol. Squasher implements the protocol’s event envelope item for errors and messages.

Keep your Sentry SDK installed as-is
Change the DSN to point at ingest.squasher.ai
Disable Sentry transaction sampling and send APM telemetry through OpenTelemetry
Errors flow to Squasher — grouped, triaged by AI, and ready in your dashboard

The DSN

Your Squasher-compatible DSN looks like this:

https://[email protected]/YOUR_PROJECT_ID

Part	Value	Where to find it
`YOUR_API_KEY`	Your Squasher API key (starts with `sq_pk_`)	Dashboard > Settings > API Keys
`YOUR_PROJECT_ID`	Your Squasher project UUID	Dashboard > Settings

Use an Ingest-only API key for the DSN. This key can only send events — it cannot read your data or manage your project. This is safe to include in client-side code, just like Sentry’s public key.

Quick Start

No package changes needed. Just update your Sentry.init():

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

Sentry.init({
  // Before: dsn: "https://[email protected]/456"
  dsn: "https://[email protected]/your-project-id",
  // Sentry transaction items are not ingested; send APM data with OTLP.
  tracesSampleRate: 0,
});

That’s it. All Sentry.captureException() and Sentry.captureMessage() calls now send to Squasher.

import sentry_sdk

sentry_sdk.init(
    # Before: dsn="https://[email protected]/456"
    dsn="https://[email protected]/your-project-id",
    # Sentry transaction items are not ingested; send APM data with OTLP.
    traces_sample_rate=0.0,
)

Works with Django, Flask, FastAPI, Celery — any Sentry Python integration.

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

Sentry.init({
  dsn: "https://[email protected]/your-project-id",
  tracesSampleRate: 0,
});

Also works with @sentry/react, @sentry/vue, @sentry/angular, and @sentry/svelte.

import "github.com/getsentry/sentry-go"

func main() {
    sentry.Init(sentry.ClientOptions{
        // Before: Dsn: "https://[email protected]/456"
        Dsn: "https://[email protected]/your-project-id",
        TracesSampleRate: 0.0,
    })
    defer sentry.Flush(2 * time.Second)
}

require 'sentry-ruby'

Sentry.init do |config|
  # Before: config.dsn = 'https://[email protected]/456'
  config.dsn = 'https://[email protected]/your-project-id'
  config.traces_sample_rate = 0.0
end

Works with Rails, Sidekiq, and any Sentry Ruby integration.

\Sentry\init([
    // Before: 'dsn' => 'https://[email protected]/456'
    'dsn' => 'https://[email protected]/your-project-id',
]);

Works with Laravel, Symfony, and plain PHP.

import io.sentry.Sentry;

Sentry.init(options -> {
    // Before: options.setDsn("https://[email protected]/456");
    options.setDsn("https://[email protected]/your-project-id");
    options.setTracesSampleRate(0.0);
});

Works with Spring Boot, Android, and any Sentry Java integration.

SentrySdk.Init(options =>
{
    // Before: options.Dsn = "https://[email protected]/456";
    options.Dsn = "https://[email protected]/your-project-id";
    options.TracesSampleRate = 0.0;
});

import * as Sentry from "@sentry/react-native";

Sentry.init({
  dsn: "https://[email protected]/your-project-id",
  tracesSampleRate: 0,
});

What Gets Captured

Squasher extracts all the error data from Sentry envelopes:

Sentry Feature	Squasher Support
`captureException()`	Full support — exception type, message, stack trace
`captureMessage()`	Full support — message text and level
Stack traces	Full support — filename, function, line/col, source context, in-app marking
Chained exceptions	Full support — cause chains preserved (most recent used for grouping)
Breadcrumbs	Preserved as bounded structured JSON, including navigation, HTTP, and console
User context	Full support — id, email, username, IP address
Request context	Full support — URL, method, headers
Tags	Object and pair-array formats; filter with `sentry.tag.<key>`
Extra data	Preserved as bounded JSON in `sentry.extra_json`
Release tracking	Full support — release version used for source map resolution
Environment	Full support — production, staging, etc.
Contexts (OS, browser, runtime, device)	Preserved as JSON and mapped to OpenTelemetry resource attributes
Exception mechanism (handled/unhandled)	Mapped to `exception.mechanism.*` attributes
SDK info	Mapped to `telemetry.sdk.name` and `telemetry.sdk.version`

Squasher applies the same kind of safety bounds as Sentry: messages, tag keys and values, breadcrumbs, stack frames, contexts, and extra data are capped before storage. This prevents a malformed event from creating unbounded ClickHouse attributes while preserving the newest breadcrumbs and both ends of long stack traces. The complete envelope body is limited to 1 MiB.

What’s Different (and Better)

Feature	Sentry	Squasher
AI triage	Paid add-on	Included on every Squasher plan
AI auto-fix PRs	Not available	Available on Team and above
Error grouping	Built-in fingerprinting	Same fingerprinting + AI context
Transactions/APM	Full APM suite	Not supported (error monitoring focused)
Session replay	Available	Available
Cron monitoring	Available	Not supported

Squasher is focused on developer-first incident debugging with AI, replay, logs, and uptime context in one workflow. If you use Sentry mainly for error monitoring and still need other products for the rest of the incident story, Squasher is built to consolidate that stack.

Supported Sentry SDK Versions

Platform	Minimum Version
JavaScript (`@sentry/node`, `@sentry/browser`, etc.)	7.0.0
Python (`sentry-sdk`)	2.0.0
Ruby (`sentry-ruby`)	4.0.0
Java / Android (`sentry-java`)	3.0.0
Cocoa / iOS / macOS (`sentry-cocoa`)	6.0.0
.NET (`Sentry`)	3.0.0
PHP (`sentry/sentry`)	4.0.0
Go (`sentry-go`)	0.1.0
React Native (`@sentry/react-native`)	3.0.0

These versions correspond to Sentry SDKs that use the envelope protocol. Older SDKs that use the deprecated /store/ endpoint are not supported.

Source Maps

If you’re using source maps with Sentry, you can upload them to Squasher using our API:

# Upload source maps for a release
curl -X POST "https://api.squasher.ai/v1/projects/YOUR_PROJECT_ID/sourcemaps" \
  -H "x-squasher-key: sq_pk_your_admin_key" \
  -F "release=your-release-version" \
  -F "files=@dist/app.js.map"

Squasher will automatically resolve minified stack traces against your uploaded source maps, just like Sentry does.

Source map uploads require an API key with sourcemaps:write permission, separate from the ingest-only key used in your DSN.

Running Both During Migration

You can send errors to both Sentry and Squasher simultaneously during migration. Most Sentry SDKs don’t support multiple DSNs natively, but you can use a tunnel or proxy approach:

Run Squasher alongside Sentry for a week to verify parity
Compare grouped issues in both dashboards
Switch the DSN once you’re confident

For an error-only Sentry setup, the DSN is the only setting you need to change. If anything goes wrong, restore the previous DSN.

FAQ

Do I need to remove the Sentry SDK?

No. You keep the Sentry SDK installed. You’re just changing where it sends data. Think of it like changing the database URL — same driver, different destination.

Will my Sentry integrations still work?

All Sentry SDK integrations (Express, Django, Flask, Rails, etc.) continue to work. They capture errors using the SDK’s hooks. Sentry performance integrations and transaction envelope items are not part of this compatibility endpoint.

What about Sentry performance monitoring?

The Sentry-compatible endpoint ingests event envelope items only; it does not ingest Sentry transaction items. Leave Sentry performance tracing disabled (tracesSampleRate: 0 in JavaScript SDKs, or the equivalent setting in other SDKs). Send traces, spans, and performance metrics to Squasher through the OpenTelemetry integration instead.

Is my API key safe in client-side code?

Yes, if you use an Ingest-only API key. This key can only send error events — it cannot read data, manage projects, or access any other resources. This is the same security model as Sentry’s public DSN key.

What if I use sentry-cli for releases?

Squasher has its own source map upload API. You’ll need to update your CI/CD pipeline to upload source maps to Squasher instead of (or in addition to) Sentry.

Agent handoff

Migrate Sentry SDK error traffic to Squasher for project <project_id>. Change the DSN or tunnel destination, disable Sentry transaction sampling, upload matching source maps for the same release, compare one known error group, and ask before removing Sentry-specific alerting or release steps.

Get Started

SDKs

Integrations

Migrations

Cloud Log Drains

Features

API Reference

Migrate from Sentry

Connect a Sentry organization

How It Works

The DSN

Quick Start

What Gets Captured

What’s Different (and Better)

Supported Sentry SDK Versions

Source Maps

Running Both During Migration

FAQ

Agent handoff

​Connect a Sentry organization

​How It Works

​The DSN

​Quick Start

​What Gets Captured

​What’s Different (and Better)

​Supported Sentry SDK Versions

​Source Maps

​Running Both During Migration

​FAQ

​Agent handoff

Connect a Sentry organization

How It Works

The DSN

Quick Start

What Gets Captured

What’s Different (and Better)

Supported Sentry SDK Versions

Source Maps

Running Both During Migration

FAQ

Agent handoff