diagnose-issues

Parallel debug agents diagnose UAT gaps by spawning one agent per defined gap, investigating root causes concurrently, and aggregating evidence into a consolidated diagnosis. This orchestration reduces overall time to diagnosis.

Parallel root-cause diagnosis is an automated process where multiple debug agents investigate UAT gaps simultaneously. It is needed when UAT.md contains multiple defined gaps and a test suite is available, requiring concurrent evidence collection to speed up root-cause identification.

No, parallel debug agents require gaps to be defined in UAT.md before invocation. The orchestrator parses this file to identify each gap and spawn one agent per gap, so missing or undefined gaps cannot be processed without prior documentation in the file.

Yes, a test suite is required to run parallel root-cause diagnosis on UAT gaps. The debug agents rely on the available test suite to investigate symptoms, collect evidence, and determine root causes for each gap defined in UAT.md during the orchestration process.

The best way to hand off diagnosed UAT gaps is to aggregate the root causes, symptoms, and artifacts collected by each parallel debug agent, then update UAT.md with precise diagnoses. This output feeds directly into plan-phase --gaps for targeted fixes.

Sequential UAT gap diagnosis is slow because each gap is investigated one at a time. Parallel debug agents address this by spawning one agent per gap to investigate root causes concurrently, aggregating findings, and reducing overall time to diagnosis.