Clone This Codex

Fast Path

Clone the repo, then create a clean starter project:

python pipeline/scripts/bootstrap_new_codex.py --project-title "New Research Codex" --primary-subject "Primary Subject" --output .\new-research-codex

The output folder gives you the framework without Steinmetz content.

What The Starter Gives You

• codex.config.json
• empty sources/source_catalog.json
• a source manifest template
• a claim-record template
• directories for raw sources, processed data, concepts, math, diagrams, glossary, comparisons, hidden gems, and public site data
• a starter charter and README

First Real Source

For the first source, create a manifest like:

{
  "source_id": "first-source",
  "collection_id": "new-research-codex",
  "title": "First Source",
  "creator": "Primary Subject",
  "year": 1900,
  "source_type": "book",
  "authority_refs": [],
  "raw_files": [],
  "rights_status": "unknown",
  "processing_status": "raw"
}

Then add the source to sources/source_catalog.json, run or adapt the extraction scripts, and generate public pages only after the source trail is visible.

Completion Discipline

A cloned codex should not say a source is decoded because it has a page. It is decoded only when the source has custody, text, structure, extracted concepts, extracted equations, extracted figures, glossary candidates, quotes, crosslinks, and visible verification status.

Recommended Adaptation Order

1. Rename the project and site title.
2. Add the first source and manifest.
3. Run source extraction.
4. Generate source readers.
5. Add concept and equation pages only after source passages exist.
6. Add diagrams and tools for the topic’s actual subject matter.
7. Add comparisons and interpretive layers last.