deep-interview

To turn vague ideas into approval-ready specifications, use a Socratic clarification loop with mathematical ambiguity scoring to iteratively expose hidden assumptions. It enforces single-question rounds until constraints and success criteria are explicitly approved before execution.

Socratic interviewing for requirements clarification is an iterative process asking targeted single questions to reduce ambiguity. It mathematically scores clarity dimensions, locking the weakest one until ambiguity drops below a configured threshold and execution is explicitly approved.

To generate specs for modifying existing code in a brownfield repository, perform repository-first exploration to gather context, then ask confirmation questions grounded in repo evidence. This topology-based handling tracks up to six components, scoring active areas independently to avoid overfitting.

Spec generation works with multi-component software requests by confirming and tracking one to six top-level components. It scores active components independently and rotates targets to prevent overfitting one area, ensuring balanced ambiguity reduction across the entire topology.

Use an ambiguity scoring execution gate for software specs when execution must not start until goals, constraints, and success criteria are fully clarified. It computes weighted clarity scores and blocks execution until ambiguity drops below your configured threshold and you approve the path.

Limitations of gating execution with ambiguity scoring include the requirement for explicit user approval at every threshold step, which prevents autonomous processing. Additionally, it restricts handling to a maximum of six top-level components, limiting complexity for highly distributed software architectures.