Modern QA2026Architecture Decisions
Log inJoin

Library Book 1 Architecture Decisions

Architecture Decisions

12.1🔒OverviewParts I through III gave you the pieces. This part assembles them into a working framework -- starting where every credible framework…96 words
12.2🔒ADR-001: Agent-Driven vs Traditional ExecutionDecision. Use an agent-as-orchestrator model: the agent reads test definitions, decides how to interact with the application through the…123 words
12.3🔒ADR-002: Playwright CLI Skill as Primary InterfaceDecision. Of three candidates -- Playwright MCP, the CLI taught via SKILL.md (playwright-cli install --skills), and the client library…137 words
12.4🔒ADR-003: Test Definition FormatDecision. Structured YAML with natural language steps:95 words
12.5🔒ADR-004: Fresh Sessions for CI, Reused Named Sessions for DevelopmentDecision. Named sessions (-s=name) persist across commands; lifecycle policy differs by environment:86 words
12.6🔒ADR-005: Snapshot-and-Screenshot DebuggingDecision. Capture screenshot + YAML snapshot on every failure. Raw DOM via eval only on request.69 words
12.7🔒ADR-006: Three-Tier Error RecoveryDecision. When a command fails, recovery escalates in cost:153 words
12.8🔒ADR-007: Test Result ReportingDecision. Three output formats -- console for CI logs, JSON for programmatic consumption, and a failure-artifact directory per failing test:40 words
12.9🔒ADR-008: Parallel Test ExecutionDecision. Sequential by default in development (one reused session). Parallel via named sessions -- each worker gets its own isolated…66 words
12.10🔒Key Takeaways- Eight ADRs form the foundation; write them down with context, decision, consequences, mitigations - Agent-as-orchestrator buys resilience…97 words
12.11🔒Q&ASelf-Assessment Quiz1. What four elements does an ADR record, and why does that matter? 2. Why did ADR-002 choose the CLI skill over MCP, and when is MCP still…178 words
12.12🔒Exercises[Beginner] Exercise 12.1: Take the YAML format from ADR-003 and write a definition for a password-reset flow, with hints expressed as…74 words
12.13🔒Career Translation- Authored eight Architecture Decision Records for an AI-driven test automation framework on the Playwright CLI, covering execution model…141 words
12.14🔒Q&AInterview Depth CheckPrompt: A team of 5 QA engineers asks whether tests should be defined in code (Playwright test files) or in YAML with natural language…297 words