Supplemental Corpus: From Code-Centric to Intent-Centric Software Engineering

## PurposeThe supplement enables reproducibility, replication, and transparency for an RTA-dominant, IPA-informed public-discourse and document analysis of GenAI and agentic systems in software engineering. ## Important evidence boundary- Peer-reviewed and archival sources support scholarly claims.- Technical preprints and benchmarks support emerging technical context.- Public talks, interviews, essays, product announcements, and X-originated posts are public-discourse corpus artifacts and are not treated as peer-reviewed evidence.- Long copyrighted transcripts and full X post text are not redistributed. ## Core files- `corpus_register.csv`: 49-source corpus with evidence-role labels and alignment notes.- `coding_matrix.csv`: source-to-code/theme traceability matrix.- `codebook.csv`: provisional semantic and latent codes.- `themes.csv`: final themes with support counts and caution notes.- `researcher_case_profiles.csv`: idiographic case profiles.- `peer_review_alignment_audit.csv`: reviewer-style audit and revisions.- `doi_reference_audit.csv`: DOI/reference validation and evidence role audit.- `manifest_sha256.csv`: file integrity manifest. ## Replication guidance- `maturity_pathway.csv`: stage-based maturity model replacing deterministic year-by-year forecasts.- `SHA256SUMS.txt`: authoritative external checksum manifest for GitHub/Zenodo package files.