A practical guide for your next QA task.
Start with the full workflow, from a changed requirement to connected tests, runs, failures, tracker updates, and release evidence. Then go deeper into each QA practice.
Run the Checkout test on Chrome and Firefox and tell me when it is done.
Done. Chrome passed 11 of 11. Firefox passed 10 of 11, and one step self-healed after a button moved. Nothing looks broken from the dev changes. Want the report?
The library
Find the workflow you need
Search by task or narrow the library to the topic you are working on.
21 guides to explore
Axon6 min read
What Axon can do across an entire release
See how one plain-English request can become connected work across requirements, tests, UI and API checks, failures, trackers, and the latest release-readiness assessment.
Read guideRelease Practice8 min read
How to tell whether a release is ready to ship
Most teams answer this from memory in a meeting. Here is how to turn it into criteria you set in advance, including the awkward middle case where you ship anyway.
Read guideRelease Practice7 min read
Release quality gates: what to actually gate on
Most gates get switched off within a quarter. The ones that survive share a few properties, and none of them is strictness.
Read guideRelease Practice7 min read
A QA sign-off template that is not a rubber stamp
Most sign-off documents record that somebody approved. The useful ones record what they knew, what they accepted, and why. Here is the difference, and a template you can copy.
Read guideImport & Export4 min read
Import your test cases from Excel or CSV
Bring your spreadsheet test cases into AxonQA in one upload. Columns map automatically, each row becomes a case with its steps and expected results, and you can export, edit, and re-import at any time.
Read guideTest Creation4 min read
Turn a user story into test cases
Import a story or write one here, and AxonQA generates happy-path, negative, and edge cases from its acceptance criteria, each traceable to the criterion it verifies.
Read guideAutomation4 min read
Record or generate automation you own
Record a real flow in the browser or generate automation from a test case grounded in an explored session. Either way you get maintainable, reviewable automation you can extend, and when you want to take it off the platform, request a migration and our team hands you the full codebase.
Read guideAxon5 min read
Run your whole regression from a sentence
Ask Axon in plain English to run your tests by description or tag, on the browsers you name. It launches in the background and reports back, so anyone on the team can run the suite.
Read guideCross-Browser4 min read
Cross-browser testing without the setup
Run the same tests across Chrome, Firefox, and WebKit in the cloud, with a separate verdict for each browser and no machines to build or maintain.
Read guideAPI Testing5 min read
Test your APIs alongside your UI
Organize requests into collections, chain them into real flows, run them against any environment with datasets, and verify the contract your interface depends on.
Read guideDiscovery5 min read
Let Axon explore your app
Point Axon at your running product to map the reachable pages, journeys, forms, and locators in AxonQA Cloud, or choose the local AxonQA Agent for a private target.
Read guideAI Grounding5 min read
Ground the AI in your real app
Feed Project Brain your documents, screenshots, and imported stories, and let app discovery map your running product, so generation uses real selectors and journeys instead of guesses.
Read guideSelf-Healing5 min read
Self-healing that repairs locator drift without hiding failures
Cheap, certain checks first and AI reasoning last. Locator drift can be repaired, assertion failures stay red, and every fix is reviewable and reversible.
Read guideRequirements4 min read
Catch requirement drift before it ships
AxonQA monitors connected Jira or Azure work items, shows the acceptance-criteria diff, and lets you generate suggested tests for the change in one click.
Read guideInsights5 min read
Track release readiness
Coverage, AI-powered dashboards, scheduled stakeholder reports, and alert rules to Slack, Microsoft Teams, or email turn scattered results into a clear answer to are we ready to ship.
Read guideSecurity5 min read
Test private or firewalled applications
The local agent runs inside your network and connects outbound only, so a one-time pairing lets you explore and test private or firewalled applications without opening a door in.
Read guideIntegrations4 min read
Bring your own AI agent
AxonQA speaks the Model Context Protocol, so assistants like Claude Desktop and VS Code, and any MCP-compatible client, can generate cases, run suites, and read results end to end.
Read guideIssue Trackers5 min read
Connect Jira or Azure DevOps and keep them in sync
Import stories and acceptance criteria from Jira or Azure DevOps, turn them into tests, and let results and status flow back, so your tracker and your tests never drift apart.
Read guideScheduling4 min read
Put your testing on a schedule
Run regression overnight, smoke every hour, and deliver a readiness report every Friday, on a cadence you set, so testing happens without anyone remembering to start it.
Read guideReporting4 min read
Turn runs into reports you can share
Build evidence-based reports from your real runs, keep them in a library, and schedule delivery, so stakeholders get the picture without logging in.
Read guideAlerts4 min read
Know the moment something breaks
Set rules for the conditions that matter, a failed regression, a drop in pass rate, a new flaky test, and get told through your chosen channels and the in-app inbox.
Read guideInside AxonQA
These practices, built into the product
The guides describe how we think about test quality. The product runs on the same ideas: tests grounded in your real requirements and flows, self-healing that escalates from quick certainty to careful AI review, and health metrics that make flakiness impossible to ignore.
- Healing escalates, cheapest and most certain checks first.
- AI is used after deterministic checks; expected results remain unchanged.
- Every fix is recorded with its reasoning, and is reviewable and reversible.
Same run, two very different causes.
The app changed, the test did not need to. Fixed during the run, and reversible.
The expected total did not match. Left red with evidence for review.
Cheap, certain checks run before anything expensive, and no fix ever rewrites what your test expects.
Get the next guide when it lands
We write about what actually works in AI-assisted testing, roughly monthly. No noise, unsubscribe anytime.
Your email, nothing else. We never share it.
Put these practices to work
Generate structured test cases from your stories, then validate them with real runs on your own app.
14-day trial · No card required · Sample project included