Ask a QA lead what actually slows a manual pass down and the answer is rarely finding bugs. It is that the third of reports written at 4pm on a Friday cannot be reproduced on Monday, and the tester who wrote them has moved on to a different build.
Most quality assurance tools respond to that by adding required fields. That reduces the number of reports without improving them, because the fields testers skip — build number, device, console state, the exact sequence of screens — are the ones a person is worst at reconstructing and a machine is best at capturing.
What TraceItX does for a manual test pass
The tester never leaves the app they are testing. They press a bubble or a keyboard shortcut, draw on the screenshot, blur anything sensitive, and send. What lands is one AI-ready envelope with the context already in it.
Nothing to install per tester
The reporter is inside your build. No extension, no account, no second tool — testers never see a TraceItX screen at all.
Unlimited testers, every tier
Seats and reports are unmetered including on Free. The meter is monthly active installs: 1,000, 10,000, 100,000 or custom.
The repro path, captured
Up to 128 breadcrumbs on one clock record the screens and actions before the report, whether or not the tester wrote them down.
Same shape on every platform
Web, iOS, Android, tvOS, Fire TV, Android TV and more return an identical typed payload. Triage rules are written once.
Redaction before the buffer
Text is scrubbed as captured, marked views are masked on-device, and the tester can blur the screenshot by hand.
A median report is about 84 KB, mostly the screenshot, so a full test pass costs very little to move or store. The SDKs are 14 KB gzipped on React, 28 KB of JavaScript on React Native, a 92 KB archive on Android and a 180 KB binary on iOS.
Start on the free tier: 1,000 monthly active installs, unlimited testers and reports, all SDKs, and AI triage on 50 reports a month. Artifacts are kept 14 days on Free and the widget carries a TraceItX badge — if your triage cycle runs longer than a fortnight, start on Starter instead.
The questions QA leads actually ask
“Does it put bugs in Jira?” Not by itself. There is no Jira app, and this is a real gap rather than a positioning nuance — BugHerd, by contrast, documents a Jira integration you connect from a project’s Integrations panel, and has rebuilt it to add status synchronisation between Jira and BugHerd columns plus user mapping between accounts. It also documents Linear and Slack integrations and commit hooks for GitHub, GitLab and Bitbucket. With TraceItX, reports arrive as signed webhooks and someone on your team writes the receiver that maps them into your tracker. Price that engineering time in.
“Our testers won’t learn another tool.” They do not open one. The reporter is a control inside the build they were already testing.
“What does it cost when twelve people file bugs?” Nothing extra. This is where per-member pricing diverges sharply: BugHerd’s published tiers cap team members at 5, 10 and 25 for $42, $67 and $125 a month respectively, and Marker.io lists Starter at $39 and Team at $149 a month per workspace. TraceItX charges for reach, not people.
“We test a native app, not a website.” Then check this first on anything you shortlist. BugHerd’s own support documentation states it is designed to work on websites and web applications. TraceItX ships native SDKs for iOS, iPadOS, tvOS, Android, Android TV and Fire TV alongside the web one.
“What about PII in screenshots and network data?” Scrubbing happens as text is captured rather than at upload, marked views and subtrees are masked on the device, and the tester blurs what they choose before sending. Storage is within the EEA where possible; retention is 14 days on Free, 90 on Starter, 12 months on Pro, each with a 14-day grace period before permanent erasure. No customer data trains models.
“How do we know a webhook is really from you?” Every delivery is signed with HMAC-SHA256 over the timestamp and raw body — the same scheme Stripe uses, so existing verification helpers translate — with retries, a stable delivery id across attempts, a dead-letter queue and a per-subscriber circuit breaker behind it.
“Can we self-host?” No. Reports transit the ingest service before reaching your receiver. That is the architecture, not a setting, so a strict data-egress policy rules it out.
Where this does not fit
Unlike broad quality assurance software suites, TraceItX handles the reporting step and nothing either side of it.
It does no bug testing of its own — it does not run tests. There is no device cloud, no browser and OS matrix, no automation runner — teams needing real-device coverage across many configurations pair it with a platform such as BrowserStack, whose free trials are documented as 30 minutes of Live, 60 minutes of Automate, 100 Screenshots and Responsive runs, and 30 minutes of App Live.
It is not an error monitor either: no alerting, no release health, no crash-free-rate dashboards, no performance metrics. Crash reporting is on by default and produces the same envelope a person would have filed, but the product’s centre of gravity is a report someone chose to send.
Two narrower limits worth knowing before a trial. On the web it supports React only — there is no framework-agnostic build, so a Vue or server-rendered site cannot use it at all, and a web quality assurance pass over a marketing site is outside what this covers. And Roku, Samsung Tizen and LG webOS need a conversation with the team rather than a package install, so “ten platforms” is true while “ten platforms you can self-serve on a Friday afternoon” is not.
If reproducibility is your bottleneck
Wire the reporter into one build before your next pass and compare a fortnight of reports against your last cycle. Unlimited seats means every tester can file without anyone buying a licence, and the free tier needs no card — the honest constraint is the 14-day artifact window, so run the comparison inside it.