What problem does it solve?
Deep Interview prevents wasted execution by converting vague or ambiguous ideas into mathematically gated, unambiguous specifications before any autonomous work begins.
Core Features & Use Cases
- Socratic ambiguity reduction: Asks one targeted question at a time to expose assumptions and clarify the weakest dimension until ambiguity falls below a configured threshold.
- Explicit ambiguity scoring: Scores goal clarity, constraint clarity, and success-criteria clarity (and context clarity for brownfield) after each answer, transparently showing progress.
- Brownfield-first code grounding: Uses a codebase exploration step before asking user confirmation questions when extending existing systems.
- Execution readiness gate: Produces a spec artifact only when clarity is sufficient, then routes into the next pipeline stage.
Common use cases include requirements discovery for complex initiatives, preventing “that’s not what I meant” outcomes in autonomous development, and safely handling incident triage when details are uncertain.
Quick Start
Run deep-interview to convert your vague idea into a ready-to-execute spec by answering the first Socratic question about what exactly should happen.
