Playwright is a framework for web testing and automation that drives Chromium, Firefox, and WebKit through one API.
- Stars
- 1,192
- Forks
- 196
- Open issues
- 44
- Last commit
- 7 days ago
What is it?
Hercules is an open-source testing agent that runs end-to-end tests written in Gherkin, the plain-English Given/When/Then format. Users describe a scenario in a feature file, and Hercules uses an LLM-driven planner and browser navigation helpers built on Playwright to carry out the steps, then reports results as JUnit XML and HTML, with video, screenshots and network logs saved as proofs.
Beyond UI flows, the project covers API testing, security scanning through Nuclei, WCAG 2.0 to 2.2 accessibility checks, and visual assertions. It supports per-agent LLM routing through an agents_llm_config.json file, including OpenAI-compatible endpoints and LiteLLM proxies, and can connect to remote browsers over CDP such as BrowserBase, BrowserStack and LambdaTest.
Hercules can be installed from PyPI, run as a Docker image, or built from source, and it can also run as an MCP server. Custom Python scripts can be invoked from Gherkin steps with access to Playwright.
Key features
- 01Runs Gherkin feature files as automated browser end-to-end tests without writing code.
- 02Covers UI, API, security (Nuclei), accessibility (WCAG 2.0 to 2.2) and visual validations.
- 03Produces JUnit XML and HTML reports, with video, screenshots and network logs as proofs.
- 04Routes planner, navigation and helper roles to different LLMs via configuration.
- 05Connects to remote browsers over CDP, including BrowserBase, BrowserStack and LambdaTest.
- 06Supports mobile device emulation using Playwright device descriptors.
Use cases
- Automating regression tests for complex web applications from plain-English scenarios.
- Running browser tests inside CI/CD pipelines using the Docker image.
- Adding security and accessibility checks to existing end-to-end test suites.
- Letting QA staff who do not code write and maintain automated UI tests.
Quick start
- 1
Install Hercules from PyPI
pip install testzeus-hercules - 2
Install Playwright and its browser dependencies
playwright install --with-deps - 3
Set the LLM_MODEL_NAME, LLM_MODEL_API_KEY and LLM_MODEL_BASE_URL environment variables for your model provider
- 4
Run Hercules on a Gherkin feature file
testzeus-hercules --input-file opt/input/test.feature --output-path opt/output --test-data-path opt/test_data
Things to know
- Licensed under AGPL-3.0, a strong copyleft license that may restrict how you can use it in proprietary products.
- Running tests requires an LLM provider API key or a self-hosted model.
Topics
FAQ
- What is testzeus-hercules?
- Open-source AI testing agent that turns Gherkin feature files into automated end-to-end browser, API, security, accessibility and visual tests. Hercules is an open-source testing agent that runs end-to-end tests written in Gherkin, the plain-English Given/When/Then format. Users describe a scenario in a feature file, and Hercules uses an LLM-driven planner and browser navigation helpers built on Playwright to carry out the steps, then reports results as JUnit XML and HTML, with video, screenshots and network logs saved as proofs.
- Is testzeus-hercules free to use? What license does it have?
- testzeus-hercules is open source under the AGPL-3.0 license. Check the license terms before using it in commercial products.
- Is testzeus-hercules still maintained?
- Actively maintained. The last commit to testzeus-hercules was on Oct 14, 2026.
- What can I use testzeus-hercules for?
- Automating regression tests for complex web applications from plain-English scenarios. Running browser tests inside CI/CD pipelines using the Docker image. Adding security and accessibility checks to existing end-to-end test suites. Letting QA staff who do not code write and maintain automated UI tests.
- How do I install testzeus-hercules?
- Install Hercules from PyPI: pip install testzeus-hercules Install Playwright and its browser dependencies: playwright install --with-deps Set the LLM_MODEL_NAME, LLM_MODEL_API_KEY and LLM_MODEL_BASE_URL environment variables for your model provider Run Hercules on a Gherkin feature file: testzeus-hercules --input-file opt/input/test.feature --output-path opt/output --test-data-path opt/test_data
- Where can I get testzeus-hercules?
- The source code and installation instructions are on GitHub: https://github.com/test-zeus-ai/testzeus-hercules
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
An MCP server and Figma plugin that let AI agents like Cursor and Claude Code read and modify Figma designs programmatically.
Code-first Python framework from Google for building, evaluating, and deploying AI agents, optimized for Gemini but model-agnostic.
Memory operating system for LLMs and AI agents that stores, retrieves, and manages long-term memory as an editable graph.
Python RL framework that trains multi-step LLM agents on their own experience using GRPO, with a client-server design that plugs into existing apps.
Enterprise-grade agentic workflow platform for building and running AI agents, with model management, MCP tool integration and RPA automation.