Files
republic-os/docs/legal-corpus.md
T
Fabio 511d65f9a0 Policy: dev mode until v1 — main is the workbench, raw bytes off-repo
This repo is a prototype testing a theory, not yet a mirror the public
relies on. The immutability disciplines (never-overwrite, never-rewrite,
raw-first-in-repo) are reframed as the v1.0 target, not day-3 law — a
mirror that force-pushes while preaching immutability would betray its
own founding principle. Until v1: history is malleable and raw source
snapshots live off-repo on the depot (.gitignore'd), with only the
product (records) and provenance (manifests + SHA-256) committed. The
147MB of already-tracked raw zips are dropped from HEAD (git rm --cached,
a forward delete — the bytes stay on the depot).

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-07-06 10:51:35 -04:00

1.8 KiB

Legal Corpus

Republic OS mirrors law as deterministic terrain before it interprets law as intelligence.

The legal pipeline is deliberately mechanical:

official source -> raw snapshot -> normalized legal tree -> manifest -> checksums -> git diff

The legal corpus is the complete United States Code — every section of all 53 titles that carry content (Title 53 is reserved and empty), 59,740 records, from the Office of the Law Revision Counsel USLM XML at release point 119-100 (current through Public Law 119-100). One source, one edition, end to end.

Pre-v1 note: the raw .zip/.xml snapshots live off-repo on the depot, not in git — only the manifests and checksums (which anchor them by SHA-256) are committed. The "raw-first, in-repo" discipline turns on at v1.0. See the README.

Layout

data/legal/raw/us/code/          (off-repo pre-v1; hashes committed below)
  title-NN/
    xml_uscNN@119-100.zip
    uscNN.xml

data/legal/manifests/
  us-code-title-NN-119-100.json

data/legal/checksums/
  us-code-title-NN-119-100.sha256

legal/us/code/
  title-NN/
    chapter-101/
      section-10101.md

The raw ZIP is the provenance anchor. The generated Markdown is the readable legal tree. The manifest records source hashes, output hashes, release metadata, and the section list.

Discipline

Legal corpus ingestion does not summarize, classify authority, infer relationships, or answer legal questions. Later layers can extract citations, definitions, amendments, and authority links from this base. The base itself is just a reproducible mirror.

Rebuild

make legal-us-code
make validate

The success condition is that rebuilding from the same pinned source produces the same normalized legal files and checksums.