skills.rest
ejoongseok
ejoongseokCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

qa-run-light

Execute QA-*.md checklist steps via MCP Playwright and generate evidence-backed reports.

8|2|Updated May 3, 2026
Search Tags:#qa#mcp#insights#regression#playwright#test execution#evidence reporting
One-click install
npx skills add https://github.com/ejoongseok/claude-settings --skill qa-run-light
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: qa-run-light
Source: https://github.com/ejoongseok/claude-settings/tree/main/claude-code/skills/qa-run-light
Command: npx skills add https://github.com/ejoongseok/claude-settings --skill qa-run-light

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you quickly execute QA-*.md checklist test steps with MCP Playwright and generate a lightweight, evidence-backed report so you can identify pass/fail/flaky/manual issues faster.

Core Features & Use Cases

  • Fast QA checklist execution (Light mode): Parses Step headers, checkbox tool calls, and expected outcomes from a single QA markdown file.
  • Evidence-first reporting: Collects screenshots and failure evidence (console, network errors, DOM snapshot) per step.
  • Actionable 7-category insights: Produces intervention-focused insights with required six fields (discovery, evidence, action, effort, confidence, priority) and P1/P2/P3 ranking.
  • Guardrails & operational controls: Supports stop-on-fail, only P1 selection, sample runs, dry-run planning, storage size checks, and blocked-manual escalation.

Quick Start

Use the qa-run-light skill to run the QA file at 'QA-MyFeature.md' once in dry-run mode to preview the steps without executing them.

Frequently Asked Questions about qa-run-light

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I execute QA markdown checklists with Playwright to collect test evidence?▼

QA markdown checklists can be executed using MCP Playwright to run tool-based checks per step, capturing screenshots, console logs, network errors, and DOM snapshots as evidence-backed execution results for pass, fail, flaky, or manual outcomes.

What is the best way to generate lightweight QA reports from E2E test steps?▼

Lightweight QA reports are generated by parsing markdown checklist files for step headers and expected outcomes, then producing actionable insights with discovery, evidence, action, effort, confidence, and priority fields ranked P1, P2, or P3.

Can I dry-run QA test steps without executing Playwright browser actions?▼

Yes, you can run QA markdown files in dry-run mode to preview parsed steps and expected outcomes without executing browser actions, allowing you to validate test plan structure and tool allowlists before actual test execution.

Does MCP Playwright support stop-on-fail and P1-only test execution modes?▼

MCP Playwright integration supports stop-on-fail execution, P1-only selection for prioritized checks, sample runs, and guarded escalation to manual handling when blocked flows require user intervention during E2E verification.

How do I structure QA markdown files for automated checklist execution?▼

QA markdown files require YAML frontmatter with name and description fields, step headers with checkbox tool calls, expected outcomes, and optional manual or pending handling tags to enable deterministic tool allowlisting and strict evidence collection.

What are the limitations of running regression tests in light mode?▼

Light mode focuses on single QA markdown file execution with evidence collection but requires strict YAML frontmatter, deterministic tool allowlisting, and manual escalation for blocked flows, making it less suited for unattended bulk regression runs.

Related Skills

Explore More Software Engineering Skills→
officebeatsofficebeats
8

test-scenarios

Generate structured test scenarios from user stories with objectives and expected outcomes.

Community
Intermediate
jovermierjovermier
1

quality-severity

Classify issues and code review findings into P1, P2, or P3 severity levels.

Community
Basic
stlwolfstlwolf
2

playwright-browser

Automate browser operations and DOM inspection via Playwright MCP.

Community
Advanced
hidonguyenhidonguyen

ck:fix

Diagnose and fix code issues with verification and regression prevention.

Community
Advanced