AI quickly drafts what takes you hours: test cases, bug reports, test data, scripts. It doesn’t know your product: what deserves testing, how severe a defect is, and the call to ship remain yours.
AI helps QA testers with tasks such as “Write a bug report,” “Write the test cases for a feature,” and “Review a spec before testing.” Start with a single task, in Claude: allow 5 minutes the first time, and always review the result before you use it.
For each task: how it goes today, then with the tool. And the time it takes the first time.
Claude · Screenshot attached to the conversation
TodayA poorly described bug comes back marked “cannot reproduce,” and you lose two days going back and forth with the developer.
With ClaudeA complete, reproducible bug report to paste into your tracking tool.
Claude · Excel file created in the conversation
TodayWriting forty test cases by hand, with the steps and the expected result, takes a day. And edge cases get dropped when the deadline tightens.
With ClaudeAn Excel file of test cases, linked to the spec criterion by criterion, ready to import.
ChatGPT · “Chat” mode
TodayThe gaps in a spec show up in the middle of testing, when every question costs a round trip with the developer and the product manager.
With ChatGPTThe list of questions to ask before development, ranked by risk.
Also
Gemini Notebook · “Chat” panel
The expected rule, with the exact passage to cite in the bug report, or confirmation that the rule isn’t written anywhere.
Wispr Flow · Hands-free mode: Fn + Space on Mac, Ctrl + Win + Space on Windows
The session log, with the exact steps for each finding, written while you were testing.
Claude · Excel file created in the conversation
A file of fictional data, with the edge cases flagged and the expected result for each.
GitHub Copilot · Agent mode, in VS Code
A readable test script, already run, that replays the flow at every release.
Claude Code · “Plan” mode
The likely causes ranked, how to confirm each one, and a clear verdict: a defect in the test or a defect in the product.
Julius · Excel or CSV file uploaded to the chat
The counts by severity and by module, the line chart of open and closed bugs, and the list of serious bugs remaining.
ChatGPT · “Chat” mode
A one-page report a committee can read, where all that’s left is to write your recommendation.
All 10 use cases on this page are in Nova, each with its steps and its prompt.
Try the interactive demoWhat to check
A screenshot taken during testing sometimes shows real customer data. Hide it before attaching. The suggested severity is a suggestion: you set it.
Write a bug reportClaude · Screenshot attached to the conversationThe prompt to copy
You’re helping me write a bug report a developer can reproduce on the first try. Here are my notes, with a screenshot attached: [paste your notes]. Environment: [version, browser or device, profile used, data set]. Write: a short title that says where and what, the preconditions, the numbered steps (one action per step), the actual result, the expected result with the rule involved ([spec reference]), the frequency, and the impact on the user. Suggest a severity and justify it. Stay factual: no guesses about the cause. At the end, list the information missing to reproduce it.
In Nova, “My path” starts from your level: one next step at a time.
The first moves, with free tools.
In your jobWrite the test cases for a feature
In your jobReview a spec before testing
In your jobFind a business rule in the specs
In your jobDictate findings during exploratory testing
For a team: everyone moves forward in their own job. You see the team’s progress, never one person’s activity.
It reads the spec and returns the test cases in an Excel file ready to import. It also turns your notes and a screenshot into a reproducible bug report.
Recommended planFree to start. Pro, $20 per month ($17 per month if billed yearly). For company documents: Team, $25 per user per month ($20 if billed yearly).
You upload the requirements document and the specs. It answers from those documents only and shows the exact passage: useful for deciding between a bug and intended behavior.
Recommended planFree: 100 notebooks, 50 sources per notebook, 50 questions per day. Included if your company is on Google Workspace.
During exploratory testing, you say what you’re doing and what you see. The finding is written into your notes or the ticket, without letting go of the mouse.
Recommended planFree, $0: 2,000 words per week on computer. Then Pro, $15 per user per month ($12 if billed yearly).
In VS Code, agent mode writes the script for an automated test from a test case written in plain English, then runs it.
Recommended planFree, $0: 2,000 code suggestions per month and limited use of agent mode. At a company: Business, $19 per user per month.
It reads the whole automated test folder and explains why a test fails one time in three. In “Plan” mode, it doesn’t touch a single file.
Recommended planPro, $20 per month ($17 per month if billed yearly): it’s the same subscription as Claude. For company code: Team, $25 per user per month ($20 if billed yearly).
You upload the bug export and get the counts by severity and by module, plus the charts for the test summary report.
Recommended planFree, $0: 25 credits per day. For weekly use: Plus, $20 per month, or $16 per month billed yearly.
Prices taken from the vendors’ official pages, dated on each tool page. These subscriptions are not included in Nova. Brands are named to identify the tools. No affiliation.
Nova recommends 6 to start with: Claude, Gemini Notebook, Wispr Flow, GitHub Copilot, Claude Code, and Julius. 5 of them can be used for free to get started.
It reads the spec and returns the test cases in an Excel file ready to import. It also turns your notes and a screenshot into a reproducible bug report. The plan to choose: Free to start. Pro, $20 per month ($17 per month if billed yearly). For company documents: Team, $25 per user per month ($20 if billed yearly).
With “Write a bug report,” in Claude (Screenshot attached to the conversation). Allow 5 minutes the first time. A complete, reproducible bug report to paste into your tracking tool.
No. Nova is software, not an accredited training provider: it issues no diploma or certification. It costs €19 per month for one person, and €129 per month for a team of up to ten people.