OSS TanbouOn this page
Overview
TestZeus Hercules is a testing framework that uses LLM agents and Playwright to execute web workflows from Gherkin scenarios. It targets UI, API, security, accessibility, and visual validation and can retain HTML/JUnit results plus screenshots, video, network logs, and agent execution artifacts.
Features and best fit
Based on official documentation; not hands-on tested · Content checked:
Key features
Execute browser workflows from Gherkin scenarios
BDD-style steps drive Playwright browser activity with agent reasoning in the execution loop instead of encoding every interaction as fixed test code.
Sources: [2]
Store results and proof artifacts per test run
The run guide describes HTML/XML results, agent logs, screenshots, videos, and network logs under separate output and proofs directories.
Sources: [3]
Extend scenarios with a Python sandbox
Custom Python scripts can access Playwright for advanced selectors, conditional workflows, reusable logic, or data processing when plain Gherkin is insufficient.
Sources: [2]
Best fit
Fits teams reducing maintenance around rapidly changing web E2E tests
It is relevant when business scenarios should remain readable as Gherkin while some browser interaction and adaptation is delegated to an agent.
Sources: [2]
Before adoption
Review test context and credentials sent to the configured LLM provider
Execution requires model credentials or per-agent LLM configuration. Tests using production-like data should explicitly review provider data flow and secret handling.
Sources: [3]
Review the telemetry changes introduced in v1.0.2
The 1.0.2 release notes mention capturing folder paths in telemetry and documenting telemetry collection. Sensitive test environments should review telemetry configuration and transmitted data.
Sources: [4]
Review AGPL-3.0 obligations for modified network deployments
Hercules is AGPL-3.0 licensed, so modified versions exposed over a network should be reviewed for corresponding source-code obligations.
Sources: [5]
Official sources
- [1]test-zeus-ai/testzeus-hercules repository(2026-10-01)
- [2]TestZeus Hercules README(2026-10-01)
- [3]TestZeus Hercules run guide(2026-10-01)
- [4]TestZeus Hercules 1.0.2 release(2026-10-01)
- [5]TestZeus Hercules AGPL-3.0 license(2026-10-01)
Supplemental curator note
Hercules replaces some selector-heavy test maintenance with agent-driven interpretation of Gherkin scenarios. Before CI adoption, review what test context reaches the configured LLM provider, browser-side effects, and the telemetry changes documented in the 1.0.2 release.
Try it in 3 steps
- 1
Install Hercules
Install the PyPI package in a supported Python 3.11–3.13 environment.
python -m pip install testzeus-hercules - 2
Install the browser runtime and prepare a test project
Install Playwright browser dependencies and create the single-test directory structure from the run guide. Place a Gherkin feature under opt/input before execution.
playwright install --with-deps && mkdir -p opt/input opt/output opt/test_data - 3
Run the test with an LLM configuration
Run the test with the selected LLM provider. In CI, inject the API key from a secret store and review handling of production credentials or sensitive test data.
testzeus-hercules --project-base=opt --llm-model <model> --llm-model-api-key <api-key>
Growth
Growth trends · Last 30 days
1,174 Stars
Trend data is still being collected.
Development activity
Last 90 days · weekly
- Commits (last 30 days)
- 0
- Open PRs
- 20
Development activity is still being collected.
Built with
Categories and tags
Categories
GitHub data
GitHub dataView detailed GitHub data
GitHub Topics
- ai
- autogen
- automation
- browser
- end-to-end-testing
- hercules
- large-action-model
- playwright
- rpa
- testing
- testzeus
- agents
Related information
Write a related articleShare a guide or use case for this OSS in Markdown. Articles are published after administrator approval.
Explore next
- LangChain147,365 Stars
3 shared tag(s) · 1 shared category(s) · Same language
compose models, tools, retrieval, and middleware into agents
Python - Crawl4AI84,635 Stars
3 shared tag(s) · 1 shared category(s) · Same language
a Python crawler that turns browser-rendered web pages into LLM-ready Markdown and JSON
Python - Khoj37,550 Stars
3 shared tag(s) · 1 shared category(s) · Same language
Combine documents, the web, local/cloud LLMs, agents, and automations in a self-hosted personal AI
Python - PipesHub3,804 Stars
3 shared tag(s) · 1 shared category(s) · Same language
preserve source-system permissions at query time while exposing enterprise knowledge to search, RAG, and MCP
Python - Lightpanda35,680 Stars
3 shared tag(s) · 1 shared category(s)
rebuild the browser runtime for AI-driven web automation
Zig - LlamaIndex52,383 Stars
3 shared tag(s) · Same language
compose ingestion, retrieval, RAG, and agents from modular packages
Python
Report incorrect information
Tell us if any listing information is incorrect or outdated.