Stacklist_Test_Cases_updated_4
The document details test cases for the STACKLIST application, specifically outlining scenarios for the "Login & Logout" module, covering successful and failed attempts, session management, SSO, and password reset flows. It also includes test cases for the "Cards" module, addressing creation from URLs or file uploads, editing, deletion, and various error conditions.
Stacklist Product Guide / Business Logic Doc
The Stacklist product guide outlines its core mental model of Cards, Stacks, and Collection Pages, detailing their hierarchical relationships and user authentication processes. It further describes the main workspace, including the sidebar's organizational and social sections, the "+" create menu, global search functionality, and user account options.
AI Agent for E2E Tests Project
This project proposes an AI agent built with Playwright and Claude to automate end-to-end website testing by simulating a user, clicking around, filling forms, and identifying broken or confusing elements. It leverages different Claude models for planning, exploring, and reviewing, combines AI judgment with automatic checks to find issues, and generates a detailed report with screenshots, descriptions, and severity for integration into ticketing systems.
QA Roadmap
QA at Stacklist currently utilizes an E2E test suite (Cypress + TypeScript) covering main flows, security, and analytics, and plans four immediate initiatives for the next four weeks to improve bug reporting, error triage, production monitoring, and quality visibility. A 3-6 month roadmap outlines 13 further initiatives, categorized by effort and type, focusing on detection, performance, documentation, analysis, compliance, and infrastructure enhancements.
Stacklist UX UI Suggestions
Stacklist UX/UI suggestions propose reconfiguring profile page menus from a single vertical list to a two-column grid, improving screen space utilization and reducing scrolling. Additionally, the document recommends implementing press-and-hold (mobile) and click-and-hold (web) gestures to facilitate easier multi-item selection across the application.
QA Monthly Report: April 27th ~ May 29th
The QA Activity Report details the period's performance, including 135 findings logged, 99% reproducibility, and all role expectations met or exceeded. It further outlines the automated web suite's growth to 153 tests, achieving 73% coverage of automatable cases, and proposes stable test identifiers while setting future goals for web and mobile testing.
AI Failure Explanations Proposal
A proposal outlines adding AI-assisted failure explanations to a nightly CI suite using Anthropic's API to automatically classify failed tests as real bugs or flaky, saving developer time from manual stack trace analysis. This system, already built and tested, is highly cost-effective with typical usage under $1 per month, includes safety measures like a hard spend cap, and ensures data privacy.
