Intent-to-Silicon: A Deterministic Ambiguity-Reduction Framework for Natural Language Software Specification Generation

Intent-to-Silicon is a research prototype that investigates ambiguity reduction as a preprocessing step for natural-language-driven software specification generation. The framework introduces a multi-stage pipeline that transforms human language inputs into structured machine-readable YAML blueprints through semantic extraction, active disambiguation, emotional intent mapping, dependency enforcement, and safe-halting mechanisms. The repository includes: Semantic Dictionary NLP Intent Engine YAML Blueprint Generator Benchmark Dataset Experimental Evaluation Framework This release contains the preprint, benchmark artifacts, and supporting implementation used in the reported experiments.