qa-explorer

Explores a running web app with Playwright and files reproducible GitHub issues for confirmed bugs.

4|Updated Oct 24, 2025
One-click install
npx skills add https://github.com/sethdavis512/iridium --skill qa-explorer-sethdavis512
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: qa-explorer
Source: https://github.com/sethdavis512/iridium/tree/main/.claude/skills/qa-explorer
Command: npx skills add https://github.com/sethdavis512/iridium --skill qa-explorer-sethdavis512

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Scripted regression tests only cover known paths, so edge-case bugs in auth, chat, notes, settings, and admin flows slip through. This Skill performs charter-driven exploratory testing against a local dev server and converts confirmed findings into structured, deduplicated GitHub issues. ## Core Features & Use Cases - Charter-based exploration: Six scoped charters (AUTH, NOTES, CHAT, SETTINGS, ADMIN, MOBILE) with step budgets, suspicious-symptom checklists, and console/pageerror/5xx listeners. - Reproduce-before-filing discipline: Every finding must reproduce twice, with the second reproduction as a minimal Playwright script embedded verbatim in the issue. - Triage and dedupe: Severity levels S1-S4, automatic dedupe against open qa-explorer-labeled issues via gh issue list, and comments on existing issues instead of duplicates. - Use Case: Ask for a "qa sweep" of the notes feature; the Skill boots the app on port 7779, creates a fresh user, probes unicode/XSS/rapid-mutation edge cases, and files an S2 issue with a runnable repro script. ## Quick Start Ask the assistant to run a qa sweep on the chat charter and file GitHub issues for any bugs it can reproduce twice.

Frequently Asked Questions about qa-explorer

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

FAQPage Schema
How do I run exploratory QA testing on my web app with Playwright?

Boot the dev server on a dedicated port with test hooks enabled, then write throwaway Playwright specs that drive the app through a charter's edge cases. Attach console, pageerror, and 5xx response listeners so every suspicious event is captured automatically.

How do I file GitHub issues for bugs found during testing?

Reproduce each finding twice, with the second reproduction as a minimal scripted Playwright repro, then create the issue with the gh CLI using a severity-tagged title and a body containing repro steps, expected vs actual behavior, and the repro script.

Can this run against a production environment?

No. The Skill explicitly forbids running against production and requires a local dev server on a dedicated port with test-only hooks like DISABLE_AUTH_RATE_LIMIT and E2E_TEST_HOOKS enabled.

How does it avoid filing duplicate bug reports?

Before filing, it lists all open issues with the qa-explorer label via gh issue list and compares root symptoms. If an open issue already covers the finding, it adds the new evidence as a comment instead of creating a duplicate.

Why are pre-hydration button clicks reported as false bugs?

Before React hydration completes, controls accept events but do nothing, which looks like a broken button. The Skill requires calling waitForHydration before interacting so these false findings are avoided.

When should I use deterministic tests instead of exploratory testing?

Use the deterministic suites in tests/ for known regression paths like responsive guardrails and visual states. Use exploratory testing to hunt past that coverage for unknown edge cases, and use a separate fix workflow to implement repairs.