hzijad/playwright-agent-skills — explained in plain English
Analysis updated 2026-05-18
Guide an AI coding agent to write reliable Playwright tests
Set up tests that start from an already logged-in session
Mock unreliable third-party network calls in end-to-end tests
Debug a flaky or failing Playwright test using traces
| hzijad/playwright-agent-skills | 00kaku/wp-rest-playground | 1ncendium/aibuster | |
|---|---|---|---|
| Stars | 5 | 5 | 5 |
| Language | — | JavaScript | Python |
| Setup difficulty | easy | hard | moderate |
| Complexity | 1/5 | 3/5 | 3/5 |
| Audience | developer | developer | ops devops |
Figures from each repo's GitHub metadata at analysis time.
Playwright Agent Skills is a set of reusable instruction files, called skills, meant to guide AI coding agents toward writing better end to end tests with Playwright, a popular browser testing tool. The problem it addresses is that AI agents, left on their own, tend to write brittle tests: clicking elements using fragile CSS class names and adding fixed waits like "pause for two seconds" instead of waiting for the actual page state to be ready. The README shows this contrast directly with a before and after example. The "before" test clicks a CSS class and waits a fixed two seconds before checking a welcome message. The "after" version instead clicks a button by its visible role and label, then uses an assertion that automatically retries until the expected heading appears, which is the pattern the Playwright team itself recommends. The project deliberately avoids being a general purpose browser automation skill. Many existing tools are built for one-off checks, like taking a single screenshot to see if a page looks right, and so they lean on shortcuts that are fine for a throwaway script but bad for code meant to live in a real test suite and run in continuous integration. Every pattern here is written with the opposite goal: producing test code a Playwright maintainer would be comfortable merging. Four skills are included, each scoped narrowly to one job. playwright-core covers everyday test writing and review. playwright-auth-state handles starting a test already logged in using saved browser session state. playwright-network-mocking replaces unreliable third party network calls with predictable mocked responses. playwright-debugging helps diagnose and fix tests that are already flaky or failing, using traces and repeated runs. Installation is a single command using the Agent Skills CLI, which the README says works with over seventy different AI coding agents, including GitHub Copilot, Codex, Cline, and Gemini CLI, not just one specific tool. The project is released under the MIT license.
A set of narrow, reusable Playwright testing skills that steer AI coding agents away from brittle selectors and hard waits.
Use freely for any purpose, including commercial use, under the MIT license.
Setup difficulty is rated easy, with roughly 5min to a first successful run.
Mainly developer.
This repo across BitVibe Labs
double-check against the repo, no cap.