Research Codex Engine

Purpose

Steinmetz Decoded should remain a serious Steinmetz archive, but its architecture is now being preserved as a reusable engine for future collections: Tesla, ether history, classical field theory, Walter Russell, Eric Dollard, concave-earth research, sacred geometry, or any other source-heavy knowledge domain.

The reusable part is not the Steinmetz prose. The reusable part is the discipline:

source custody before interpretation;
page maps before quotations;
concepts, equations, diagrams, glossary, quotes, and hidden gems as structured data;
source fact, modern explanation, math, history, and interpretation kept in separate layers;
public readers, verification queues, and release levels that make uncertainty visible.

Core Data Model

Every future collection should preserve this minimum claim shape:

{
  "person": "Charles Proteus Steinmetz",
  "source": "Theory and Calculation of Electric Circuits",
  "chapter": "Chapter 3",
  "page": 42,
  "claim_type": "explicit_source_claim",
  "interpretation_layer": "source",
  "confidence": "needs-verification"
}

That is the rule that lets one platform contain many thinkers without blending their claims together.

Standard Pipeline

Add source files and manifests.
Extract or OCR text.
Split into chapters, lectures, pages, or sections.
Extract equations, figures, concepts, glossary terms, quotes, and hidden gems.
Generate readers, workbench pages, concordances, formula maps, visual maps, and verification queues.
Promote only reviewed claims into canonical explanation pages.
Keep interpretive synthesis labeled as interpretation, never as source fact.

Reusable Starter

The repo now includes a dependency-free bootstrap script:

python pipeline/scripts/bootstrap_new_codex.py --project-title "Tesla Decoded" --primary-subject "Nikola Tesla" --output .\tesla-decoded

It creates a clean starter workspace with source directories, manifests, templates, and source-isolation rules, but no Steinmetz-specific content.

Templates live in:

templates/research-codex-engine/codex.config.template.json
templates/research-codex-engine/source_manifest.template.json
templates/research-codex-engine/claim_record.template.json

What Not To Reuse Blindly

The Steinmetz sidebar, concept list, equation pages, source names, and interpretive examples are specific to this archive. A Tesla or ether-history project should reuse the engine and standards, then rebuild its source catalog, concepts, diagrams, and comparisons from its own material.

Long-Term Umbrella Pattern

The eventual platform can be one public research library with separate collections:

/steinmetz
/tesla
/ether-history
/walter-russell
/eric-dollard
/concave-earth
/sacred-geometry

Global concept pages can compare thinkers only if each claim remains attributed and confidence-labeled.