AI Copilot & IDE Tools — Frequently Asked Questions
What is the AI Copilot?
A chat assistant inside QAEverest that can do the work as well as talk about it. Ask from the search box in the top bar. It can generate test cases, build a suite, start a run and answer questions about a past run, showing its steps as it works.
It also answers quality questions from your own data: whether a release is ready and what is blocking it, which tests to run for a GitHub pull request, a pasted diff or a list of changed files, which tests are flaky and which failures keep recurring, and which high-risk requirements have no tests.
Will the Copilot spend my credits without asking?
No. Actions that spend credits ask you to confirm first, and the card says what will happen. Generating shows the price per test case. The cards for building a suite and starting a run don't show a credit figure.
Answering questions from your own data uses no credits and starts no work. Apart from the conversation itself, the one thing it saves is a test selection analysis, kept in your Smart Selection history.
Which editors does QAEverest work with?
- Visual Studio Code — the QAEverest extension on the Visual Studio Code Marketplace.
- JetBrains IDEs — the QAEverest plugin on the JetBrains Marketplace, for IntelliJ IDEA, PyCharm, WebStorm and the others.
- Claude Code — QAEverest's own plugin marketplace: run /plugin marketplace add abhilasha3-ai/qaeverest-claude-plugin, then /plugin install qaeverest@qaeverest.
- Cursor, Windsurf and other MCP editors — the MCP server on npm, @qaeverestsupport.ai/qaeverest-mcp, added with a few lines of configuration and your API key.
All of them run the same QAEverest server, so the QA tools are the same in each. Only sign-in differs: the VS Code and JetBrains extensions hold your key, so they don't offer the connect and disconnect tools.
What can I do from my editor?
Index and search a project, rank its testable features and generate tests that cite real code, generate functional, API, UI and mobile cases from a story or spec, run a suite and save its report, run security and performance checks and, with a personal API key, run a saved AI agent suite against its release gate.
When you ask through the QAEverest chat, generating tests, building a suite or starting a run waits for your approval. In an MCP editor, your AI agent passes that approval on for you. Direct tools such as qae_generate_testcases and qae_run_suite have no approval step: they spend credits as soon as the agent calls them.
What does the editor integration send to QAEverest?
It sends data only when you use a tool or the QAEverest chat panel: your API key, and the content you pass in, such as a story, a spec, a URL or an attached file. Tools that work on your code also send project files: indexing a project sends its source files in chunks to build a search index, and building an inventory or generating from features sends the files each feature came from. During a chat, the Copilot can also ask to read a file or search your project, and what it finds is sent to QAEverest. Scanning a project reads it locally and sends nothing.
It does not run in the background, scan your machine outside the project or collect telemetry.
How do editor tools sign in and get billed?
With a QAEverest API key that starts with qae_, or through browser sign-in where the editor offers it. A personal key spends your account's credits. A key from the API client dashboard, browser sign-in or an IDE sign-in spends that API client's balance. The qae_status tool shows your key type and what is left.
What is FindMyLocator?
A free extension for Chrome. Click an element for a unique CSS selector and XPath, plus Playwright locators by role, label, placeholder or test id where one finds that element alone. Test any locator, collect up to 100 elements into a Playwright, Selenium or Cypress Page Object, and pick inside open shadow DOM.
Record a flow and press Open in QAEverest to make it a UI test case. With your QAEverest API key, the Test cases tab writes Functional, System, Integration, Security or Performance cases from the page, a story or the recording, to copy or download, and Check text underlines spelling and grammar mistakes on the page.