
from team-attention
Systematically QA test applications (web, native, CLI) with modes for browser, computer, and CLI; produces reports, health scores, and fixes.
Systematically QA test any application — web apps, native macOS apps, Electron apps, CLI tools, interactive REPLs, or anything on screen. The skill auto-selects interaction mode (browser/computer/cli) or accepts overrides and supports three tiers: quick, standard, and exhaustive. It produces before/after health scores, detailed issue reports with screenshots, and can make minimal atomic fixes in source when code is available. Phases include planning, orientation, exploration & documentation, triage, fix loop, final QA, and report generation.
QA skill for systematically testing web, native, and CLI applications with three modes (browser, computer, CLI). No bundled scripts — SKILL.md-only skill with well-structured 8-phase workflow from test planning through fix verification. References external mode-specific files for progressive disclosure. Clean security posture with no scripts to audit, though git commit operations are assumed trusted.
Well-crafted skill with thoughtful self-regulation (WTF-likelihood heuristic, hard cap on fixes). Health score rubric is detailed and useful. No scripts to execute or audit — static analysis only.