Retrace Kit logo Retrace Kit
Sign up

One incident for every runtime outage

Related crashes cluster into a single incident. Triage the outage once, not five open tickets.

Start free — 100k events

No credit card required

Read setup guide

Related crashes go in. One incident comes out.

  1. 01 // ERRORS ARRIVE

    Events, sessions, and breadcrumbs stream in. Related crashes share a route, release, or visit.

    Signals correlating

  2. 02 // INCIDENT OPENS

    Different fingerprints on the same surface become one incident. That is the unit of work.

    Incident active

  3. 03 // YOU TRIAGE ONCE

    Investigate the outage as a whole. Resolve the incident when the cluster is handled.

    Outage closed

The incident is the unit of work

An issue is one fingerprint — the same crash type counted many times. An incident is several different issues that share a surface, so you triage one outage instead of a pile of tickets. Retrace Kit opens incidents automatically from normal SDK context — no extra config.

How incidents work →
SHP-41

TypeError: Cannot read 'id' of undefined

214 events

SHP-42

TypeError: Failed to fetch /api/pay

89 events

SHP-55

RangeError: Invalid currency code

31 events

INC-184 Active High

/checkout

One outage across cart, payment, and currency

Issues
0
Users
0
Sessions
0
  1. Candidate
  2. Active
  3. Investigating
  4. Resolved

Issues are the building blocks of an incident

Identical crashes fingerprint into one issue by message and stack. Several issues on the same surface make up the incident you triage.

Aggregation →
Aggregation in the dashboard 1,024 events → 1 issue
Events 247 today
  • 14:02:11 TypeError: Cannot read 'id' of undefined

    checkout.ts:47

  • 14:02:09 TypeError: Cannot read 'id' of undefined

    checkout.ts:47

  • 14:01:58 TypeError: Cannot read 'id' of undefined

    checkout.ts:47

  • 14:01:44 TypeError: Cannot read 'id' of undefined

    checkout.ts:47

  • 14:01:12 TypeError: Cannot read 'id' of undefined

    checkout.ts:47

+ 997 identical events hidden

Issue Open

TypeError: Cannot read 'id' of undefined

1,024
events
·
847
users
·
checkout.ts:47
Occurrences · 7d ↑ spike
Error-free users 94.2%

AI accompanies every incident and issue

Each brief covers the technical state, business impact, and fix options. It refreshes when new events change the picture.

AI brief SHP-42 high confidence

TypeError: Cannot read properties of undefined (reading 'id')

Stack

handleCheckout · checkout.ts:47

Breadcrumbs

restoreSession → /checkout → submit

Release

v2.4.1 · production · ck-prod

  1. 01

    Summary

    Checkout crashes when handleCheckout reads .id on a missing cart item after a stale session restore.

  2. 02

    Likely impact

    847 users blocked on /checkout. Conversion drops until the null guard ships.

  3. 03

    Probable fix

    Guard cart item before read and refresh session state in CheckoutForm.tsx.

Refreshes when new events change the issue context

Live in minutes, not hours

Install, init, ship. Crashes group into issues; related ones on the same route, release, or session become an incident you can triage.

~2 KB

gzipped bundle

2 lines

to go live

0

config files required

src/main.ts

main.ts package.json
// two lines. that's it.
import { init } from '@retrace-kit/sdk';
 
init({ apiKey: 'your-project-api-key' })
TypeScript UTF-8 Ln 4, Col 42

RetraceKit vs other platforms: same errors, less noise

RetraceKit runs an incident engine that turns related crashes into one place to triage.

Other platforms Typical

  • Millions of logs: You dig through thousands of duplicates of the same error by hand.

  • Overloaded UI: Dozens of APM charts and tabs you never open.

  • Complex setup: Hours spent writing mapping rules and filters.

Result: Focus shifts from writing code to managing the tracker itself.

Ideal for product teams
RetraceKit Only what matters

  • Incident engine: An automatic engine correlates related crashes into one incident ready to triage.

  • Clean tracking: Stack trace, environment, and affected users. Nothing else.

  • Zero-config: One line of code at startup. No custom configs.

Result: You see what broke, fix it, and ship.

Everything you need to triage

Built for indie SaaS and small teams who triage outages as incidents — not a raw event feed or a data lake.

Explore docs →
~ triage

Incidents

Correlate related fingerprints into one outage. Triage the cluster, not a pile of tickets.

  • Route
  • Release
  • Session
~ triage

AI summary

On every incident and issue — from stack, breadcrumbs, and release context. Read the fix before you open the repo.

  • Summary
  • Impact
  • Fix
~ metric

Error-free users

See what share of sessions ran clean today. Simpler than a stability score.

  • EFU
  • Sessions
~ context

Breadcrumb trail

User clicks, navigation, console. Replay the path before the crash.

  • Debug
  • Timeline
~ source

Stack traces

Frames, messages, release tags. Enough to open the right file.

  • JS
  • Source maps
~ notify

Team alerts

Telegram and Bitrix24. Practical channels for small teams.

  • Telegram
  • Bitrix24

Ready to keep projects stable?

100,000 events free every month. First errors and incidents in under three minutes. No credit card, no platform tax.

Create free account View pricing