AI-powered QA studio that generates and executes test cases. Turn manual text, Figma designs, Jira issues, screenshots, and recorded browser sessions into structured, traceable test cases — then auto-run them as sandboxed Playwright tests with pass/fail results and screenshots. A single-process FastAPI app with a SQLite store and a vanilla-JS SPA; works with OpenAI, Anthropic, or Google Gemini.
- Multi-source generation — manual text, Figma, Jira, screenshots (vision), and browser recordings, combinable in one request via a plugin parser architecture.
- Auto-execute — turn any test case into runnable Playwright code, run it against your base URL, and get pass/fail/error with a screenshot and console output. Generated code is cached, editable, and re-runnable.
- Projects → features → test cases — persisted in SQLite with full CRUD, project-wide dedup, source traceability, stats, and generation history.
- Multi-LLM — OpenAI, Anthropic, and Gemini with automatic provider fallback. API keys are stored in the database, never in
.env. - Export — Excel, CSV, JSON, or TestRail-style CSV, with feature / priority / search filters.
- UI — project gallery, sidebar workspace (Overview, Test Cases, Dashboard, Settings), Chart.js dashboard, dark mode, responsive.
- Python 3.11+
- At least one LLM API key: OpenAI, Anthropic, and/or Google Gemini
start.sh installs dependencies, builds the CSS, and starts the server — you only create the venv and add a key once:
git clone https://github.com/aswanirs23/AutoQA-Studio.git
cd AutoQA-Studio
python3 -m venv .venv # one-time
cp .env.example .env # then add at least one LLM API key
./start.sh # runs on http://localhost:8080Open http://localhost:8080/. On re-runs just ./start.sh again.
On Windows or if you prefer explicit steps, see manual setup.
- Non-secret config (provider defaults, DB path, auth) goes in
.env— copy it from .env.example. - API keys (LLM providers, Figma, Jira) are entered in the Settings UI and stored in SQLite;
.envvalues for these are intentionally ignored.
Full environment-variable reference: docs/configuration.md.
- Configuration — environment variables, manual / Windows setup
- Deployment — production config, Tailwind build, Docker, hosting notes
- API reference — REST endpoints and request shapes
- Input parsers — built-in parsers and how to add one
- Browser session recording — record a flow, then generate tests
- Auto-execute — run test cases as Playwright; security boundaries
- Architecture — project layout
- Known limitations & roadmap — what it doesn't do yet, and proposed fixes
Released under the MIT License © Aswani.