chenjigang4167/testhub_platform

TestHub – AI‑powered full‑stack test management platform

What it is – TestHub is an open‑source web application that helps QA teams plan, write, run and report on tests. It adds AI‑driven helpers (requirement‑to‑test‑case generation, an integrated chatbot, and a Browser‑use mode that lets a language model drive a browser) on top of a conventional test‑management stack.

Key capabilities

Area	What TestHub provides
AI assistance	• Upload a requirements document (PDF/Word/TXT) and let supported LLMs (DeepSeek, OpenAI, Anthropic, Gemini, etc.) automatically extract business requirements and generate test cases. • Dify‑based chatbot for on‑demand testing advice. • Browser‑use mode: the LLM interprets page structure (text or visual) and creates Selenium/Playwright scripts to exercise the UI automatically.
Test‑case lifecycle	Create, edit, version, archive, tag, and organise cases by project/version. Supports multi‑person review with templates, checklists and status tracking.
API testing	HTTP & WebSocket support, hierarchical collection of requests, variable substitution, assertions, scheduled runs, email/Webhook notifications, Allure‑style reports.
Web UI automation	Selenium & Playwright back‑ends, multi‑browser (Chrome/Firefox/Edge), element library, Page‑Object Model, visual script editor, video/screenshots, cron‑based scheduling, plus the AI‑driven “smart mode”.
Mobile (Android) automation	Airtest image‑recognition engine, local emulator or remote device handling, resource locking, component library, UI‑Flow orchestration, Celery‑backed async execution with real‑time WebSocket progress.
Performance testing	Locust‑based load‑test jobs with result aggregation and reporting.
Data factory	>50 utilities (random data, encoding, encryption, JSON helpers, cron expression parser, etc.) that can be referenced from API or UI tests.
Security & collaboration	JWT double‑token with auto‑refresh, blacklist on logout, role‑based permissions, multi‑project isolation, unified notifications (email, enterprise‑WeChat, DingTalk, Feishu).

Architecture

• Backend: Django 4.2 + DRF, MySQL, Redis (optional for Celery/WebSocket), JWT auth, Celery + APScheduler for async jobs, Channels/Daphne for WebSocket, Allure for reporting, LangChain‑OpenAI & browser-use for LLM integration.
• Frontend: Vue 3 + Vite, Element Plus UI library, Pinia store, Vue‑Router, Axios, ECharts, Monaco editor, i18n support.

Getting started (as described in the README)

1. Clone the repo and create a Python 3.12 virtual environment.
2. Install Python dependencies from requirements.txt.
3. Copy .env.example to .env and fill in MySQL, Redis, email and LLM API keys.
4. Create the MySQL database, run Django migrations, and create a super‑user.
5. (Optional) Load built‑in locator strategies and Android component packs with the provided management commands.
6. Start the services:
   • python manage.py runserver (HTTP API) or daphne … for WebSocket support.
   • celery -A backend worker -l info for async tasks.
   • python manage.py run_all_scheduled_tasks for cron jobs.
7. In a separate terminal, build the frontend:

   cd frontend
   npm install
   npm run dev   # dev server at http://localhost:3000
   

8. Access the UI (http://localhost:3000), the API docs (/api/docs/), and the Django admin (/admin/).

Docker users can run ./build-and-push.sh followed by docker‑compose up -d for a one‑click full stack.

Who would use it

• QA teams that want a single portal for requirement analysis, test‑case authoring, review, and execution.
• Organizations looking to augment manual testing with LLM‑generated cases or AI‑guided UI automation.
• Developers needing an extensible platform (Django apps, Vue components) to plug in custom test‑type integrations.

License – GPL 3.0 (see LICENSE).

---

All information above is taken directly from the repository’s README; no additional features have been inferred.