deep-interview

To gather requirements for vague software features, you can use a structured Socratic interview that asks one question at a time to surface hidden assumptions and eliminate ambiguity before execution. This ensures all dimensions of the request are defined.

Writing specifications for high-risk architecture decisions is best handled through a Socratic requirement elicitation process that tracks ontology and scores real-time ambiguity. This generates a clear spec file once your ambiguity drops below a configured threshold.

To handle ambiguity when integrating with a brownfield codebase, a one-question-at-a-time interview process applies context tracking to surface hidden assumptions. This structured elicitation ensures complex integration requirements are fully defined before any code execution.

Yes, you can pause a structured requirements gathering interview because the process features resume-friendly state persistence. This allows you to stop the Socratic elicitation and continue the exact session later without losing your ambiguity scoring progress.

Ambiguity scoring during requirements elicitation works by evaluating your answers across multiple dimensions in real-time. The interview continues asking focused questions one at a time until your overall ambiguity score drops below a configurable threshold.

Generated specification files are automatically written to the .omc/specs/ directory as markdown files after a requirements interview. Specifically, the spec is saved at .omc/specs/deep-interview-{slug}.md for integration with the omc pipeline.