quality-debug

What problem does it solve?

This Skill accelerates root-cause diagnosis and fix validation for software bugs by converting symptom descriptions or UAT gap lists into a structured, two-wave CSV pipeline that coordinates parallel investigative and fix agents, removing manual triage and repetitive verification work.

Core Features & Use Cases

• Wave-based CSV orchestration: Generate hypothesis rows and pre-created fix rows in a master tasks.csv, run investigation and fix waves in parallel, and merge results into results.csv and context.md.
• Automated agent spawning and concurrency control: Uses a spawn_agents_on_csv pattern with configurable concurrency and runtime limits to scale investigations and fix attempts safely.
• Discovery board & UAT integration: Append-only discoveries.ndjson for shared findings, optional --from-uat mode to import UAT gaps, and rules to update UAT and issue records after diagnosis.
• Use Case: Turn a brief bug description or a UAT phase into a session that generates 3–5 hypotheses per issue, investigates them in parallel, and attempts fixes for confirmed causes.

Quick Start

Start a CSV-wave debugging session by running quality-debug with a concise bug description and optional flags such as --yes, --concurrency, or --from-uat to control automation and parallelism.