deep-dive

To investigate an auth token expiring early, you can use a trace and interview pipeline to perform deep problem-solving. This approach investigates code, configuration, and system behavior to identify the root cause and crystallize the findings into actionable requirements.

The best way to crystallize vague requirements for brownfield analysis is a two-stage pipeline that traces existing system behavior and then injects those findings into a deep-interview. This ensures context continuity for accurate requirement generation.

No, Claude team mode is required to execute parallel traces and manage state persistence for system investigation. This mode enables the pipeline to maintain trace and spec continuity throughout the deep-interview process.

The deep-dive pipeline supports both greenfield and brownfield analysis for requirement crystallization. It uses a customizable execution path, allowing flexible standalone investigation or ralplan to autopilot execution for new and existing systems.

Trace context flows into the interview stage for problem-solving through a 3-point injection mechanism. This directly integrates the findings from the trace stage into the deep-interview, ensuring full context continuity for accurate requirement crystallization.