sj-qa

What problem does it solve?

Traditional QA processes in AI-driven development workflows often rely on biased developer self-assessments, leading to missed defects and inconsistent validation of implemented features against original requirements. This Skill eliminates that gap by enabling fully independent verification that only references original requirement documents and actual changed files, excluding developer self-evaluations to ensure unbiased results.

Core Features & Use Cases

• Independent Verification: Excludes developer self-summaries per Judge independence principle, only referencing original PM briefs and actual changed files for unbiased assessment.
• Standardized Verdict Generation: Produces clear PASS/FAIL/CONDITIONAL judgments with stored state, integrated with project progress tracking.
• Post-Deployment Monitoring: Includes canary mode for production status checks and benchmark mode for Core Web Vitals and load time performance measurement.
• Use Case: After a developer implements a new payment processing feature, use this Skill to independently verify all completion conditions from the original PM brief, run edge case tests, and generate an unbiased verdict without being influenced by the developer's self-evaluation, then monitor production performance after deployment.

Quick Start

Invoke the sj-qa skill to run an independent verification of the recently implemented user authentication feature against the original product requirement brief and receive an unbiased PASS/FAIL/CONDITIONAL verdict.