Automatic incident clustering
Related issues are grouped into one incident when they share route, release, or session context, which helps reduce scattered triage across separate tickets.
RetraceKit is incident tracking for JavaScript and TypeScript teams. It clusters related crashes into automatic incidents so small teams can triage outages from a quieter inbox without setting up a full APM.

RetraceKit is an incident tracking product for JavaScript and TypeScript apps. It groups related error fingerprints into automatic incidents so teams can treat several crash patterns as one outage when they share a route, release, or session.
The product is positioned for indie hackers, solo founders, and small teams that want a lightweight incident inbox rather than a broader application performance platform. It uses a small SDK, automatic capture of uncaught errors, and built-in context to surface incidents quickly with minimal setup.
Related issues are grouped into one incident when they share route, release, or session context, which helps reduce scattered triage across separate tickets.
Individual crashes are tracked as issues, while incidents collect multiple issues tied to the same outage surface, so teams can keep both the detail and the operational view.
Each incident and issue includes an AI-generated summary, giving teams a brief description without needing a separate summarization workflow.
The SDK is described as about 2 KB gzipped and initializes with automatic capture of uncaught errors, keeping setup lightweight.
Team alerts and third-party integrations are included, and the product notes Telegram and Bitrix24 integrations as live today.
The pricing page shows free and paid plans with different retention windows and concurrent stored event limits, while error collection continues even after limits change.
When several error fingerprints appear on the same route, RetraceKit can gather them into one incident so a team can investigate the outage as a single work item.
If a release introduces multiple new failures, the release-based correlation rule can surface them together for faster review after deployment.
When one user session triggers a chain of different issues, the session view helps teams understand the full failure path instead of separate error records.
Indie hackers and small JS/TS teams can use it as an incident inbox for crash triage without adopting a wider APM stack.
Teams that want fewer duplicate alerts can rely on issue aggregation and automatic incidents to reduce repeated noise from the same outage.
An issue is one fingerprint, meaning the same crash type counted many times. An incident is a cluster of several issues that share a route, release, or session, so the team can triage one outage instead of a pile of separate tickets.
No separate incident configuration is required. The product uses normal SDK context, including the page URL in browsers, sessionId from init, and release when you set it.
Yes. The pricing and feature pages state that automatic incidents and AI summaries are available on Free, Plus, and Pro.
The pricing page says users can email from the app or use the contact link on the page, and the team activates the plan manually and bills monthly.
It is designed for crash triage and automatic incidents for small JS/TS apps, but the site says it does not replace tools like Sentry or Bugsnag when you need session replay, distributed tracing, performance spans, or a broader APM workflow.