Changelog – what's new in QA.tech.

API and MCP for test cases, knowledge and rules, plus enforceable MFA

• –Enforce multi-factor authentication for your organizationHighlight
• –Export test case steps, and manage test cases through the APIHighlight
• –Manage knowledge items and rules through the APIHighlight
• –Chat agent can generate downloadable files

Chat agent lists runs and creates scenarios on the fly

• –Move test cases into a new scenario from chat
• –Chat agent can now list and filter recent runs

Cross-test PR verdicts, mobile build numbers, and per-run outbound IP

• –PR review verdicts now weigh all of a PR's tests togetherHighlight
• –Search your entire Jira or Linear backlog from chat
• –Clearer CORS failures in network logs
• –Mobile build numbers on results, and downloadable preview builds
• –Every test run now shows its outbound IP
• –PR review chats now stay in one conversation

Tracker comments in chat and versioned CI mobile builds

• –Chat agent reads issue comments from Jira and Linear
• –Custom version labels for API-uploaded mobile builds

PR review self-repair and a ready-made rules library

• –PR review repairs fixable test failures before shipping a verdictHighlight
• –Browse a library of ready-made PR review rules

Diff-anchored PR verdicts, cross-bot coverage, and faster dashboard

• –PR review verdicts now require evidence the diff caused the failureHighlight
• –Faster page loads across the dashboard
• –PR review testing covers mobile and tablet without manual setup
• –PR review bot cross-checks other review bots' findings

Full-file PR reviews, UUID shortcuts, and new docs

• –Copy organization and project UUIDs from the command palette
• –New docs: controlling parallel test execution
• –PoC setup checklist on docs.qa.tech
• –PR review now covers all changed files in large pull requests

PR reviews now catch coverage gaps before posting

The PR review agent runs a structured self-check after testing and fixes gaps it finds – then surfaces what it couldn't reach.

Preview-environment search and sharper PR review targeting

• –PR review targets the right path and rejects tests that don't cover the changeHighlight
• –Chat agent can now list and search preview environments
• –Follow a PR review live as it runs
• –PR review now reads the full PR conversation

Live PR review status, session-replay speedup, and mobile IP allowlist

• –Faster tests on apps that use session-replay toolsHighlight
• –Chat agent picks the right login config for your app
• –Clearer waiting messages when a review is starting
• –Mobile testing IP ranges now in docs and settings
• –"Only with reviews" now includes in-progress reviews
• –PR retros now include a plain-language summary and regression coverage
• –In-progress states and conversation links on PR review pages
• –Pull requests list now shows live review status

Unverified PR verdict, PR detail context, and device-preset autofill

• –PR reviews now surface an "unverified" outcomeHighlight
• –Allowlisting @qatech.email for email inbox tests
• –Chat messages appear immediately when you send them
• –Device presets: region selection auto-fills locale and timezone
• –Draft test cases now appear in search results
• –More context on your pull request detail page
• –Shared steps now documented under Core Concepts

MCP tools for failing tests, PR list redesign, and check-run controls

• –Investigate failing tests through the API and MCP toolsHighlight
• –Automatic reruns now reach large failing runs
• –Inconclusive PR reviews now resolve as a passing GitHub check
• –Pull Requests page redesigned as a table
• –Skip the GitHub check run for PR reviews

Mobile GPS location, chat-driven device presets, and PR page pagination

• –Set a GPS location for mobile device presetsHighlight
• –Agents identify list and table items by content, not position
• –Archive test plans through the chat assistant
• –Change-review Action now updates the GitHub PR check
• –Console and network logs visible on the final assessment step
• –Custom Fields documented, and configs explained as variables
• –Device preset egress regions now available in chat
• –Device preset settings visible before enabling mobile testing
• –PR Review check unblocks GitHub auto-merge
• –Pagination and PR number search on the Pull Requests page

Config descriptions, smarter API keys, fast-agent defaults, and one-time-use data

• –One-time-use test data configsHighlight
• –API key management improvements
• –Config descriptions help the assistant pick the right login
• –New tests default to the faster agent
• –View code diffs inline in the chat

Post-run AI analysis, geo-routing, and self-describing MCP tools

• –Post-run AI analysis and automatic rerunsHighlight
• –Bitrise CI/CD integration guide
• –Test from specific countries using geo-location routing
• –MCP tools now describe return shapes and next steps
• –PR review bot posts new GitHub comments for status updates

PR review GitHub checks, org-wide API keys, and feature search from coding tools

• –PR reviews now create a GitHub checkHighlight
• –Disable network log capture in mobile device presets
• –One API key for all your projects
• –PR review override flow, live progress, and check-run fixes
• –Project IDs are now shown and copyable in the dashboard
• –Search your app's features from AI coding tools

Project creation API, redesigned PR picker, and custom-field configs

• –Claude Desktop removed from MCP install options
• –Create projects via the API
• –Pass custom key-value data to tests via configs
• –Redesigned GitHub and GitLab PR picker in chat
• –Status badge links now filter to the right test plan

Test environments from chat, deployment-aware PR reviews, and org-scoped API keys

• –Create test environments from chat
• –Org-scoped API keys
• –PR reviews wait for your deployment before running

Rerun from chat and inline single PR reviews

• –PR reviews appear inline when there's only one
• –Rerun and retry tests directly from chat

Chat assistant edits configurations and scenarios

• –Edit test case scenarios from chat
• –Update existing configurations from chat

List runs via MCP, typed tool schemas, mobile test cases, and API key labels

• –PR review bot keeps one comment per pull requestHighlight
• –Test Cases page works on mobileHighlight
• –Agents can list runs via MCP and API
• –Chat assistant explains when file downloads aren't possible
• –MCP tools now expose typed schemas for all parameters
• –Name your API keys
• –Network settings shows the correct outbound IPs URL
• –Personal API keys
• –Tighter PR review status comment handling
• –Sign-up page is faster
• –Simpler sign-up when arriving from a shared link

Mobile overlay fixes and a clearer PR reviews list

• –Mobile session overlays and report screenshots now line up
• –PR reviews list shows what's pending and how each review started

PR reviews page, chat agent upgrades, and mobile network logs

• –PR reviews now have a dedicated pageHighlight
• –API-triggered reviews work without a mapped integration
• –Network logs for Appetize mobile tests
• –Chat agent applies test case and device preset edits immediately
• –Chat agent can inspect pull request details
• –Dismissing a pending PR review from the chat now works
• –Fewer incorrect failures when steps auto-complete
• –Old chat sessions no longer fail to load
• –PR review bot handles previously-unseen pull requests

Sign-in improvements and faster PR reviews

• –PR review bot responds immediately and handles longer runs
• –Sign-in page remembers your last-used auth method
• –Sign-in page guides new teammates through shared project links

Teams, smarter chat runs, and visual element matching

• –Access control and session security hardened
• –API-created conversations attributed to API key owner
• –Change review shows an error when the page can't be fetched
• –Chat now runs suggested tests in a single run
• –Test sessions no longer missing from chat
• –Issues weren't being created on test runs – fixed
• –Onboarding error fixed
• –Organization domains with DNS verification and auto-join
• –Outbound IP list updated
• –Auth-gated preview deployment URLs now work correctly
• –Teams management in organization settings
• –Agent finds UI elements using visual descriptions

MCP for AI coding tools, smarter onboarding, and chat-attached tests

• –MCP – Connect AI coding tools to QA.tech with one commandHighlight
• –Agent selector is now top-level; crawl trigger removed
• –Android testing now defaults to Pixel 9 Pro / Android 15
• –Captcha detection and solving improved
• –GitLab integration settings now explain the @mention trigger
• –Live status cards for background agent operations
• –Onboarding chat now researches first, then proposes
• –Open any test case directly in chat
• –PR review summary redesigned with linked test names
• –Project knowledge lookup rebuilt
• –Slack run notifications now show trigger context and results
• –Small fixes

Chat-triggered runs link back to the conversation

Run results started by the chat agent now show which conversation triggered them, with a link back to it.

Onboarding, run-trace, and verdict improvements

• –Chat-edited tests are tracked in the revisions sheet
• –Failed network requests show up in the run trace
• –A smarter onboarding chat
• –Smarter verdicts on partial test runs

More reliable login automation

• –Agent handles MFA logins more reliably
• –Faster, more reliable captcha handling

Test plans, dependencies, and PR routing

• –Cross-environment dependencies in multi-environment runs
• –Issue exports include the test steps
• –Smarter PR deployment routing
• –Test plans filter by application

Background sub-agents and richer Jira tickets

• –Better Jira tickets
• –Sub-agents run in the background

QA.tech CLI is generally available

A real command-line interface for CI, scripts, and coding agents.

Smarter error classification and Coverage polish

• –Coverage page polish
• –Smarter HTTP error classification

Coverage is available to all users

See what your application has, what's tested, and where the gaps are.

Code-aware PR reviews powered by the Assessment Agent

• –The Assessment Agent now drives PR reviews
• –Marketing site content routing
• –PR reviews now read your code

Cursor plugin and step regeneration

• –Cursor plugin
• –Regenerate all test steps in one click

Vercel auth handling and PR re-run commands

• –Re-run from the PR comment
• –Vercel-protected sites no longer fail on the first 401

New onboarding

Sign up with GitHub or Google, connect a repo, and start testing in one flow.

Test verdicts you can trust

Every test run now gets a structured verdict with reasoning, not just a pass/fail bit.

"Fix in chat" without hunting for a Run ID

Click Fix in chat from a failed test and the conversation just starts.

Chat model picker, fresh re-runs, and spending limits

• –Choose your chat model
• –"Re-run all" creates a fresh run
• –Spending limits and budget alerts

Mobile testing follow-ups

Three small but visible mobile fixes that landed together.

Configurable session state lifetime

Output state validity is no longer hardcoded to six hours.

Editing a test no longer stops an active run

Teammates can edit a test case while a run is in progress without killing the run.

Mobile live view, navigation, and Insights polish

• –Drafts and Archived navigation
• –Mobile live view with touch overlays
• –"Time saved" reads as days plus hours

Yearly subscriptions: correct metered window

Overage billing for yearly customers is now calculated against the monthly window, not the yearly period start.

Chat integrations and run-control fixes

• –Linear and Jira tool in chat
• –Stop a test while steps are being generated
• –Useful errors when a test plan ID is invalid

Spending limits with enforcement

Budget alerts now block runs when you hit your limit, not just notify.

Salesforce-heavy apps render correctly

A Salesforce-specific HTML processing fix unblocks testing on Salesforce-backed UIs.

Test Verdict Agent in production

The first wave of users started getting structured test verdicts.

Mobile testing: device presets, frames, and providers

Run tests on iOS and Android devices through Appetize and AWS Device Farm, with realistic device frames in the tracer.

Mobile reliability and issue flagging foundation

• –Issue flagging foundation
• –Mobile reliability and accuracy upgrades

Smart crawl and default test plans for new projects

• –New projects ship with Regression and Smoke test plans
• –Smart crawl on application creation

Network and console signals for evaluation

The Assessment Agent can now read network logs and console errors when judging a run.

Test ownership and clearer archive copy

• –Archive dialog reflects reality
• –Test ownership

Filter archived tests in scenarios

Archived test cases are hidden by default in scenario views, with a toggle to show them.

Larger uploads and bulk select

• –Bulk-select all test cases from the dashboard
• –File uploads larger than 50MB

Chat agent creates test plans

Ask the agent to set up a plan and add tests, without leaving the conversation.

Test creation and re-run fixes

• –"Re-run failed" works from the run results page
• –Test cases require a goal

Test plan filtering and chat-driven membership

• –Add and remove tests from a plan in chat
• –Filter and group test plans by app and platform

Cache improvements for high-volume customers

Faster, cheaper runs for accounts running thousands of tests per month.

Mobile testing in early access

The first version of mobile testing – iOS and Android – opens to early-access customers.

Why did this run error?

ERROR runs are no longer a black box.

Multi-turn evaluation foundations

Evaluation went from a single-shot decision to a real reasoning loop.

Email attachments and self-curating knowledge

• –Chat agent manages its own knowledge
• –Email attachments in tests

Dependency-aware re-runs and plan-aware chat

• –Chat is test-plan-aware
• –Re-run with dependencies

Faster runs and bulk plan management

• –Bulk add and remove from test plans
• –Faster runs via parallel tool calls

Test on draft PRs

New GitHub integration setting to opt into running tests on draft pull requests.

Test plan page redesign

Four tabs, per-schedule device preset and environment overrides.

Longer browser-agent runs and bulk test-case editing

• –Browser agent supports much longer runs
• –Edit Multiple Test Cases at Once

Bulk test editing and browser agent speedup

• –Batch-edit test case settings
• –Major browser agent speedup

Richer PR comments and a new test-cases API

• –Create test cases via the public API
• –Failure reasons in PR review comments

Test cases open in their own edit page

The side-sheet pattern is replaced with proper editor pages.

Get Email Notifications When Test Runs Complete

You can now get email notifications when test runs complete - delivered to your inbox with pass/fail status, failed test details, and direct links to results. How It Works Email notifications: Select team members from yo

Use AI Chat to Bulk-Edit Your Tests

Editing tests one-by-one is slow. Now you can describe changes across your entire test suite in chat. The AI finds relevant tests, proposes edits for each, and you review and apply them all at once. What You Can Do Fix a

Claude Haiku 4.5 Is Now the Default AI Model for All New Tests

We switched the default AI agent from Claude Sonnet to Claude Haiku 4.5. Tests run faster. Existing tests are unaffected. What Changed New default: New tests use Claude Haiku 4.5 unless you select a different agent Exist

Send Test Notifications to Microsoft Teams Channels

If your team uses Microsoft Teams, you can now receive test run notifications directly in your Teams channels - no need to switch between tools to stay informed about test results. How It Works Simple webhook setup: Crea

Tests Now Run Faster by Default

Tests now run faster by default. Agent Cache stores AI reasoning decisions from successful test runs and reuses them when your pages haven't changed, making subsequent runs significantly faster without any configuration

View Downloaded Files Directly in Test Sessions

During test execution, downloaded files appear in the Downloads dialog, but until now, you couldn't see what was inside without manual inspection. Now you can preview PDFs, images, spreadsheets, and code files directly i

Send Test Notifications to Different Slack Channels

Now you can override Slack notification channels per test run via API, with optional filtering. Different test runs deserve different audiences: CI/CD pipelines might need a deployment channel, while nightly tests go to

Agents Can Call APIs During Test Execution

QA.tech agents can now make HTTP requests during test execution to fetch test data, authenticate, validate responses, or trigger webhooks. How It Works API configs: Create an API Call config in Project Settings → Configs

Agents Can Copy and Paste Values Between Test Steps

Agents can now copy values from one step and paste them into another, making it easier to test workflows that require transferring data between form fields or pages. How It Works No setup required: Clipboard support work

Improved Getting Started Experience

New users automatically get a specialized onboarding agent that analyzes the website, detects authentication needs, and generates critical tests focused on revenue-driving flows. How It Works When you create a new projec

Automatic Exploratory Testing on Pull Requests

Code reviews usually can't catch UI bugs or broken user flows - so issues slip through even when the code looks correct. We added automatic exploratory testing that runs when preview deployments are ready, catching issue

Redesigned Dashboard with Clear Metrics and AI Chat Access

Checking test health used to require navigating multiple pages and piecing together information. Now we added key metrics directly on your project home page and created a dedicated Insights page for deeper analytics, mak

Create New Tests with AI Chat

Want to test a new feature or check if something works? Instead of writing test cases manually, just ask the chat assistant - "Explore the checkout flow" or "Test the new search feature" - and watch it creating test case

Programmatic test control API and SSH-tunnel testing

• –Create Tests via API
• –Added SSH Tunnel Proxy to Access Dev or Staging Environments Under Firewalls

Transparent Usage and Billing Metrics

Without visibility into which projects consume the most testing resources, it's hard to optimize costs or plan budgets effectively. Now you can see test case executions broken down by project and time period, making it e

Complete Network Request Inspection

Debugging API issues with limited visibility into request and response data wastes time guessing what went wrong - so we added full network inspection to the test tracer. How It Works Full Request and Response Bodies: Vi

Multiple applications per project and faster environment switching

• –Test Multiple Applications from One Project
• –Switch Between Environments Effortlessly

Automated Testing for Preview Deployments

Catching issues after code is merged to production is too late—so we added automatic testing for preview environments on pull requests. How It Works Automatic Detection: QA.tech detects preview deployments from Vercel, N

Smart Rerun: Only Test What Failed

Rerunning entire test suites to catch flaky tests or verify fixes wastes time and resources—so we made the rerun button smarter. How It Works Selective Rerun: The rerun button now intelligently identifies and reruns only

AI That Understands Your Application

Generic AI suggestions can miss critical context about your specific application and workflows—so we added a knowledge system that helps the AI understand what makes your project unique. How It Works Upload Documentation

Organize Tests with Sub-Scenarios

As test suites grow, finding and managing related tests becomes increasingly difficult—so we added hierarchical organization to help you structure tests logically. How It Works Nested Scenario Structure: Create parent an

Visual Context: Images and PDFs in Chat

Explaining complex bugs or design requirements through text alone can be time-consuming and prone to misunderstanding—so we added support for visual attachments in chat. How It Works Upload Screenshots and Mockups: Attac

Test Creation Through Natural Conversation

Creating and managing tests often means navigating through multiple screens and remembering where everything is—so we built a conversational interface that lets you work naturally. How It Works Natural Language Test Crea

Keep Your Workspace Organized with Project Archives

Managing multiple projects over time can lead to a cluttered workspace—so we added a simple way to keep things organized without losing your history. How It Works Archive Inactive Projects: Easily archive projects you're

Bug Reproduction Made Effortless

Manually reproducing issues slows everyone down—so we re-imagined the process, making it effortless and delightfully fast. How It Works Simply describe the bug—like "Incorrect phone number in sign up doesn't give a valid

E2E Testing Across Every Screen Size

Consistent Testing Across Every Screen Size Our customers told us testing applications across various device screens can be cumbersome—so we simplified it. Introducing Device Presets , a streamlined way to manage device-

Tracer view, GitHub status badges, and dependency graph

• –Enhanced Tracer: View Page and Agent Data
• –GitHub Status Badges
• –Preview Test Changes in Edit View
• –Run with Dependencies in Test Edit View
• –Test Dependency Graph

Improved Network and Console Logging

When debugging test automation, tracking events across a full test session can be overwhelming, especially when identifying exactly which actions correspond to each step. With our improved network and console logging, yo

Highlight Interactive Elements on the Page

It's frustrating when your automated tests fail because the AI agent misses elements it should interact with. To solve this, we've introduced a simple way to visually confirm exactly what elements your agent sees and can

Version History for Tests

Introduced a new feature that tracks all changes made to test cases over time—test steps and configurations. Teams can now review who made changes, when they were made, and what was changed. This addition ensures a compl

AI-Suggested Test Cases

AI-Suggested Test Cases QA.tech now intelligently suggests valuable test cases by automatically analyzing your website. Our AI-powered agent can: Log into your website (if authentication details are provided). Crawl your

Test Execution Insights

QA.tech now provides insights into your test executions. With clear visual analytics, you'll instantly understand your test performance and trends: What's Included: Test Execution Trend : Easily track the number of tests

Easier debugging, file downloads, and project contexts

• –Easier Debugging
• –Automatically Test and Verify File Downloads
• –Project Contexts

Quick Summaries for Failed Tests

Quick Summaries for Failed Tests, making it much easier to diagnose and understand test failures at a glance. How it Works: Hover to View Details : Quickly see why a test failed by hovering over the status in the test li

Test Plans

Test Plans allow you to organize and manage groups of test cases that can be executed together as a single unit. Overview Test Plans are collections of test cases that you want to run together regularly. This feature pro

File Uploads with Custom Files

We've enhanced file upload testing capabilities to support both default and custom files, making it easier to test different file import scenarios. How it Works: Default Test Files : QA.tech provides built-in test files,

Multiple Configs for User Role Testing and More

We've introduced the ability to define multiple configurations for test runs, allowing for more flexibility in how tests are executed. Now, you can create and manage different test setups based on user roles, file upload

Jira integration and mouse event support

• –Jira Integration
• –Mouse Events

Trello Integration

We've made it easier to track and manage bugs with our new Trello integration. Now, you can send bug reports directly to your Trello boards, streamlining collaboration and ensuring issues are addressed efficiently. How i

Stop a test from the editor and AI step suggestions

• –Stop Test in Edit View
• –Suggestions for Test Steps

Duplicate tests easily

We're introducing Test Duplication , a simple way to duplicate any existing test. This feature is ideal for creating variations and saving time during test setup. How it works: Open the Test Menu: Choose an existing test

Edit scenarios

We've added the ability to edit test scenarios , giving you full control over how your tests are organized. Scenarios group related tests, making it easier to keep everything structured and aligned with your goals. What

Instantly add detected tests

We're excited to announce a significant improvement to our Detected Tests feature. Now, you have the power to instantly add detected tests , eliminating the need to wait for the tests to be manually created by us. How it

Add your own custom tests

We're excited to introduce the possibility to add your own custom tests. Simply describe the feature you're building, and our AI agent will generate a tailored test plan. Here's how it works Describe your feature: Provid

See and Request Tests

Our new Detected Tests feature shows you a list of tests that we have detected based on our analysis of your website or web application. You can then request these tests. Detected Tests We're excited to introduce Detecte

File Upload Tests

We have implemented a new feature that allows you to upload files in our tests, expanding the scope of testing to cover even more complex user journeys. File Uploads It is important to verify that file uploads not only a