Does it support sharded test runs?

Absolutely. It automatically detects if reports need merging and runs the Playwright merge-reports utility to consolidate data from multiple shards.

Does it require any specific CLI tools?

To download artifacts directly from GitHub, the GitHub CLI (gh) must be installed and authenticated on your system.

Can it detect backend or API errors?

Yes, it extracts and analyzes Playwright traces to find failed GraphQL mutations and network requests that may have caused the UI to fail.

How do I provide the CI report to the skill?

You can provide a GitHub Actions run URL, a PR URL, a local ZIP file, or a folder path containing Playwright blob reports.

How does it help with fixing the actual code?

It categorizes failures, checks historical fix attempts, and provides deep context to specialized sub-agents to implement automated code repairs.

Playwright Failure Analyzer

Name: Playwright Failure Analyzer
Author: saleor

bysaleor

•

980

•

Security & Testing

Diagnoses and fixes Playwright E2E test failures by analyzing CI reports, screenshots, and GraphQL network traces.

This skill streamlines the debugging process for Playwright end-to-end tests by automatically fetching, merging, and parsing failure reports from GitHub Actions or local files. It intelligently identifies root causes by examining DOM snapshots, accessibility trees, and GraphQL network traffic within Playwright traces, helping developers distinguish between application bugs and flaky tests while providing actionable insights for sub-agents to implement fixes. It specifically excels at identifying failed GraphQL mutations and transient UI issues that are often missed during manual triaging.

Key Features

01Historical fix attempt tracking to prevent repetitive or failed repair strategies

02Deep trace analysis including GraphQL mutation and query error detection

03Comprehensive report merging and parsing for distributed CI shards

04Automated GitHub Actions artifact downloading via Run or PR URLs

05Visual context extraction from screenshots and accessibility trees at the point of failure

06980 GitHub stars

Use Cases

01Debugging failing CI pipelines in a Playwright-based E2E environment

02Investigating flaky tests by analyzing network responses and transient UI states

03Automating the triage of large-scale test failures across multiple distributed shards

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add saleor/saleor-dashboard analyze-failures

For use in Claude.ai and ChatGPT

Download Skill