audit-reference-originality

Build a registry of all supplied references, inventory the current site's text, brands, numbers, images, assets, videos, and structure, then compare category by category. Run the included Python helper to find exact file matches, text overlap, and reused numbers as leads for manual review.

Use the build_evidence_inventory.py script with --site and --reference arguments. It computes SHA-256 hashes for exact file matches, eight-word shingle overlap for text, shared number tokens, and Git-history blob matches for deleted or renamed copies.

Yes. The audit inspects renamed, deleted, and replaced files using git log, git show, and git log -S. The helper script also hashes reference files and matches them against all historical Git blobs to surface past exposure separately from current output.

No. Findings are reported as originality risks, overlaps, or red flags with severity levels, never as legal conclusions. Copyright, trademark, license, and fair-use questions require legal counsel; the audit only documents observable evidence.

Generic patterns like black backgrounds, large sans-serif type, standard pricing tables, fade-ins, common icon libraries, or shared frameworks are not red flags alone. Escalation requires distinctive combinations, direct file evidence, or reproduced identity, copy, or composition.