A QATonic product
Qaira
AI QA engineer in your terminal
What is Qaira?
Qaira is QATonic's AI QA engineer in your terminal. Start with a feature description or an OpenAPI specification to design tests, generate Playwright code, run a suite, classify failures, and review locator repairs.
Workflow history stays in a local hash-chained audit log that you can verify for tampering. Qaira uses your chosen Anthropic or OpenAI provider with your own API key. Local history does not mean all processing is offline: AI workflows send prompts and selected context to that provider.
How to get started
- 1
Install and check your environment
Use Node.js 22.13 or newer, install qaira from npm, and configure your model-provider API key in your environment. Install Playwright and its browsers in the project for browser workflows.
- 2
Initialize and design a test
Run qaira init inside your project, then qaira doctor. Describe a feature with qaira test, or supply an OpenAPI specification for API testing. Inspect the test design and generated code before running it.
- 3
Run, review, and verify history
Run tests, inspect failure classifications, and preview locator repairs before applying them. Review AI decisions and use qaira history --verify to check the local history's hash chain.
npm install -g qaira
qaira init
qaira doctor
qaira test "user can log in with email and password"
qaira history --verifyBefore you use it
Qaira is pre-1.0. AI output and locator repairs need review, and model-provider usage may incur charges. Browser workflows require Playwright; accessibility scans also require @axe-core/playwright. The explore command produces an exploration charter and unverified hypotheses from a page snapshot, rather than a full autonomous browser session. Tamper-evident history detects changes; it does not make the local log immutable.
Questions about Qaira
How do I install Qaira?
Run npm install -g qaira with Node.js 22.13 or newer. Configure your provider API key, then use qaira init and qaira doctor inside your project.
Does Qaira test both web UI and APIs?
Yes. Qaira supports Playwright-based web testing and API test generation from an OpenAPI specification. Check the npm README for each command's current requirements.
Is Qaira completely offline?
No. Its audit history is local, but AI workflows use your configured model provider. Review the package's data-handling documentation before sharing project context.
How is Qaira different from QAPrompt?
QAPrompt is a reusable QA resource library. Qaira is an executable CLI that uses a model provider and testing tools to help design, generate, execute, and maintain tests.