deep-interview

To turn vague ideas into executable specs, you can run a Socratic interview workflow that asks targeted questions and mathematically scores ambiguity to generate unambiguous, testable execution specifications with clear acceptance criteria.

To clarify requirements for a brownfield codebase, use a requirement discovery process that explores the existing repository first and cites codebase evidence to confirm how the new feature should fit before generating a final spec.

Socratic questioning prevents premature implementation by asking one question at a time and applying ambiguity gating, refusing to proceed with spec generation until clarity scores for goals, constraints, and success criteria drop below a configured threshold.

Yes, you can generate requirement specifications for greenfield projects without writing code first by using a Socratic interview to extract ontology, track convergence, and output a ready-to-run spec document to your project directory.

When crystallizing ambiguous development tasks, the output format is a markdown specification file written to .omc/specs/deep-interview-{slug}.md, which includes structured acceptance criteria and an ontology convergence summary for downstream execution.