Current Project Status
Last audited from local files on 2026-06-28.
This page records the final child-repository project snapshot. The repository is now inactive and retained for source history, citation context, and audit provenance. New implementation work, publication updates, support-rating changes, reliability design, and claim promotion now belong in ashitaka-emishi/political-pathology-workbench.
Summary
The project is no longer just a scaffold. The expanded pre-v1 corpus boundary is frozen as of 2026-06-28. It contains six cases and 41 manifest documents: Lincoln (5), American Revolution (11), Napoleon (11), Hitler (8), French Revolution (2), and British World War I (4). All six cases are rights-reviewed, acquired where rights permit, normalized, segmented, and pass corpus verification.
The metaphor workflow now has a strict MIPVU identification layer before CMT/Koenigsberg interpretation.
The main remaining gap before publication-ready reuse in PPW is reliability and claim promotion rather than corpus construction. The now-frozen expansion added targeted American Revolution, Lincoln, and Napoleon documents plus draft French Revolution and British World War I corpora, then ran them through first-pass Codex-assisted MIPVU review, CMT/Koenigsberg annotation where supported, concordance, case analysis, cross-case synthesis, validation, and traceability regeneration. These annotation-forward artifacts remain provisional until independent human review. Lincoln retains complete model-reliability artifacts for the prior three-document core and designed human-reliability cohorts, but required human coder submissions are not complete and the expanded Lincoln documents are not included in that reliability design yet. American Revolution, Napoleon, Hitler, French Revolution, and British World War I model and human reliability workflows remain absent or incomplete. Public support ratings and promoted analysis claims remain draft until reliability, historical citation, and traceability gates are satisfied.
Case State
| Area | Current state |
|---|---|
| Lincoln | 5 manifest documents; corpus verification passes 5/5; all 16,198 lexical units have Codex-assisted first-pass MIPVU decisions; all 5 documents have generated annotation artifacts. The prior 3-document core has complete historical model-reliability artifacts and designed human reliability cohorts, but required primary coder submissions are not complete and reliability resampling is still needed for the expanded corpus. |
| American Revolution | 11 manifest documents; corpus verification passes 11/11; all 34,890 lexical units have Codex-assisted first-pass MIPVU decisions; 4 documents have generated annotation artifacts. Jefferson’s rough draft and Washington’s Newburgh Address have reviewed MIPVU worklists but no CMT annotations from this conservative pass because no metaphor or uncertain units were selected. |
| Napoleon | 11 manifest documents; corpus verification passes 11/11; all 23,224 lexical units have Codex-assisted first-pass source-language MIPVU decisions; 9 documents have generated annotation artifacts. The pinned Gallica/Plon 1796 Army of Italy proclamation has reviewed MIPVU worklists but no CMT annotations from this conservative pass because no metaphor or uncertain units were selected. |
| French Revolution | 2 manifest documents; corpus verification passes 2/2; all 17,123 lexical units have Codex-assisted first-pass MIPVU decisions; both documents have generated annotation artifacts. No support rating or reliability sample exists yet. |
| British World War I | 4 manifest documents; corpus verification passes 4/4; all 18,213 lexical units have Codex-assisted first-pass MIPVU decisions; all 4 documents have generated annotation artifacts. No support rating or reliability sample exists yet. |
| Hitler | 8 manifest documents; corpus verification passes 8/8; normalized, segmented, and all 86,752 lexical units have Codex-assisted first-pass source-language German MIPVU decisions. Source-derived corpus artifacts remain gitignored under fair-use constraints. Results remain provisional until independent human review. |
| Cross-case | Synthesis, mapping, and validation scaffolds exist under cases/x-case/; comparative protocol and moral-equivalence guardrails are defined, but findings remain draft until case-level ratings, reliability, and claim traceability are promoted. |
Frozen Corpus Boundary
The original working corpus was the verified document set for the four initial historical cases:
- Lincoln: 3 documents.
- American Revolution: 9 documents.
- Napoleon: 10 bulletins.
- Hitler: 8 documents.
That rule was superseded by GitHub milestone 7, Pre-v1 corpus expansion window, tracked by issue #174. The expansion window is now closed for corpus growth. The frozen pre-v1 corpus is:
- Lincoln: 5 documents.
- American Revolution: 11 documents.
- Napoleon: 11 documents.
- Hitler: 8 documents.
- French Revolution: 2 documents.
- British World War I: 4 documents.
No further cases or documents should be added before v1 except to repair a document-specific defect that blocks verification or materially corrupts an already admitted source. Wholesale redownload, starter-corpus replacement, and opportunistic source growth remain out of scope. The selection rubric for new cases and targeted document additions is recorded in case-selection.qmd; the ranked candidate matrix is recorded in docs/corpus/pre-v1-expansion-candidate-matrix.md; rights and provenance decisions for proposed expansion sources are recorded in docs/corpus/pre-v1-expansion-rights-provenance-review.md. Deferred candidates, including Imperial Japan, Wilson WWI, Stalin WWII, Mao/CCP, Napoleon’s Farewell to the Old Guard, Saint-Just source additions, and Hitler support material, are future-work candidates rather than part of the frozen pre-v1 corpus.
Checks Run
Most recent local checks run on 2026-06-28:
npm run status
npm run validate
npm run test:model-reliability
npm run pipeline
quarto render
npm run site:table-overflowThese passed in the current expansion window. JSON validation reported that authorized local Hitler artifacts match committed integrity hashes.
Previously recorded corpus checks:
Corpus verification passes for all six cases:
python3 scripts/verify-corpus.py --case lincoln
python3 scripts/verify-corpus.py --case am-rev
python3 scripts/verify-corpus.py --case napoleon
python3 scripts/verify-corpus.py --case hitler
python3 scripts/verify-corpus.py --case fr-rev
python3 scripts/verify-corpus.py --case wwi-britainPreviously recorded JSON validation:
python3 scripts/validate-json.pyPreviously recorded MIPVU worklist generation for all four historical cases:
python3 scripts/generate-mipvu-worklist.py --case lincoln
python3 scripts/generate-mipvu-worklist.py --case am-rev
python3 scripts/generate-mipvu-worklist.py --case napoleon
python3 scripts/generate-mipvu-worklist.py --case hitlerFrozen-Project Boundaries
- Do not add new cases, documents, reliability runs, support ratings, or publication claims in this child repository.
- Treat unresolved reliability, human-review, citation, traceability, and claim-promotion work as PPW work.
- Use this repository only for provenance-preserving maintenance such as broken-link repair, citation/source-history clarification, or lifecycle notices.
- Keep the published Quarto site available as a reference snapshot, not as the active publication home.
- Promote only validated, traceable claims through PPW review gates.
Status Page Generation
scripts/generate-project-status.py generates project-status.qmd from the status JSON files and corpus artifact counts:
python3 scripts/generate-project-status.pyThe generated page is useful for local status inspection. This hand-maintained page records the audit notes and unresolved decisions that the generator cannot infer.