QAEverest ships as an MCP server and wraps it in a listing for every place you already work — the npm registry, the VS Code and JetBrains marketplaces, the Claude Code plugin catalog, Atlassian Jira and GitHub. Each one is free to install, connects with a short device code instead of a pasted key, and bills straight to your own QAEverest account.
The listings differ only in how they sign you in. Behind every one of them sits the same qaeverest-mcp server calling the same public API, so a suite generated in Cursor is the same suite your teammate sees in Jira.
qaeverest-mcpThe stdio MCP server every other listing wraps. Install it globally and register it in any MCP-compatible client — Claude Code, Cursor, Windsurf — to expose the full QAEverest tool set to your agent.
qaeverest.qaeverest-mcp“QAEverest — AI Test Case Generation” registers the MCP server for Copilot Chat and any MCP-capable agent in the editor, so test generation happens beside the code it covers.
ai.qaeverest.jetbrainsA deliberately thin onboarding plugin: Tools → QAEverest → Sign In runs the device flow, stores your key in the IDE PasswordSafe — never plain settings — and hands off to AI Assistant’s MCP client.
qaeverestA plugin catalog you can install from directly, wrapping qaeverest-mcp through .mcp.json and adding first-class slash commands to the Claude Code session.
QAEverest Test GeneratorAn issue panel and issue-context module that reads only the summary and description of the issue you open it on, generates test cases, and pushes the ones you approve back as sub-tasks.
PR-native QA botBring QAEverest into the pull request itself — branch and PR diffs import as ready-to-test stories, and the QA bot comments risk analysis on the changes under review.
Every extension uses the same device-authorization flow. You never see a credential in your clipboard, and the key that reaches your editor is minted fresh for that device.
qae_ key (or a v1 JWT exchanged from it) signs service calls. A separate portal token, valid for eight hours, handles account work — regenerating keys, billing and approving devices.QAEVEREST_API_KEY in the environment — the same variable every listing reads.Once the server is registered, these arrive in your agent’s tool list. Cost is metered per unit against your credit balance and written to your usage log — checking the balance itself is free.
generate_testcasesUser story → functional, negative, edge and accessibility test cases1generate_api_testsREST story or spec → an API test suite2generate_ui_testsUI test generation, returned as an async job you poll5generate_mobile_testsMobile test generation, returned as an async job you poll5security_scanOWASP, security-header and SSL/TLS scan of a URL10performance_testLoad, stress, spike or soak test of an endpoint10usageRemaining credits, request quota and current per-call ratesfreeNothing is charged through a marketplace. Usage lands on the QAEverest account you control, so the same balance covers the web app, the IDE and the API no matter which listing spent it.
Type a user story the way you would inside your editor and preview what the MCP tool call hands back to your agent — before you install anything.