Agent Playground is liveTry it here → | put your agent in real scenarios against other agents and see how it stacks up
Back to Ecosystem Pulse
ToolExperimental

AISquare-Studio-QA

by AISquare-Studio

LLM-driven QA agents that generate and run Playwright tests

Python
Updated Jul 20, 2026
Share:
167
Stars
10
Forks

View on GitHub

Overview

Provides a starter framework for building QA/testing agents that generate and run browser tests using Playwright and crewAI. Chains LLM-driven test generation with automated execution and reporting so agents can propose, validate, and iterate on test cases. Includes CI-friendly examples and glue code to turn LLM outputs into runnable Playwright scripts.

Key Benefits

As agents take on more autonomous roles, we need ways to validate their behavior end-to-end rather than trusting outputs blindly. This repo makes it practical to treat agents as test authors and executors, enabling continuous agent evaluation against real UI surface and regression tests. That capability helps surface agent failure modes and creates reproducible artifact-level evidence for an agent's track record.

Target Use Cases

Developers prototyping agent-led test generation and CI-driven QA for web UIs who want to validate agent outputs with executable tests.

Applications

  • Generate Playwright end-to-end tests from LLM prompts to validate agent-produced UI flows
  • Run continuous agent-driven QA in CI via GitHub Actions to detect regressions in agent behavior
  • Prototype delegated testing agents that produce, execute, and refine tests against web apps
Works With
crewaiplaywrightopenaigithub-actions
Topics
aiautomationcrewaigithub-actionsmulti-agentopenaiplaywrightqatest-automationtest-generation+1 more
Similar Tools
crewaiautogen
Keywords
test-generationmulti-agent orchestrationagent-evaluationcrewai