# NATHAN AI 🚀 — All Interfaces & Build Recovery (additive atlas)

Parent: **NATHAN AI → TABOR ROOT → TABOR 1-2-3 FACTORY → ONE / independent products**.

## Continue into the Program Jobs workbench

The separate [Program Jobs and Completion Atlas](../program-jobs/) combines the older 92-name historical recovery catalog with current code-bearing directories, HTML entrypoints, architecture modules and archive anchors. It provides jobs, tasks and five evidence gates, plus an offline local inventory matcher. Its 357 overlapping evidence records are **not** 357 independent completed products.

## What this update actually adds

- A **separate**, searchable recovery and navigation interface at app-factory/nathan-ai/build-recovery/index.html.
- A source-indexed registry.json with **76 existing repository HTML paths** as of the integration branch's starting main snapshot. The recovery interface itself is **new** and therefore not counted among those original 76. All 74 paths are source-file locations, not verified live web deployments.
- An additive list of **92 named system/reference records**, clearly marked recovered references. These do not mean 92 new engines were coded, trained, manufactured or deployed.
- **14 owner-held/historical archive anchors**, identified by filename and description without publishing the owners' private source contents.
- A live read of the existing **73-title Games & Puzzles Atlas** in ../games/catalog.json, preserving its original development stages and without copying or changing that original game catalog.
- Owner-selected source-file session indexing (text preview/search for selected readable files up to 8 MiB, exact download from original File objects). No automatic upload, GitHub publication, IndexedDB or localStorage writes; browser tab navigation/refresh clears chosen files.
- Cross-links from the existing Nathan AI main hub and ONE home, plus a direct listing in the main hub's existing directory.

## Historical recovery vs current source

registry.json includes historical recovery labels from earlier archives: 6,056 master registry rows; 383 selected technical records (273 normalized titles); roughly 5,447 saved file records, which include versions/copies. **These counts are not refreshed repository facts** and are not counts of completed apps, full books, games, or deployable binaries.

The old broader recovery index named Nathan_AI_Recovery_Index_2026-11-01.html is owner-held; the full private archive was not copied into the public code. The offline v073 package and its original modules are likewise **not** asserted present in this GitHub repository; only a source anchor/reference exists here.

Existing Library Mind has a separately published source reader, original book plans and actual language profiles. The separate **33,000,000 address library goal**, **133,333,333 master blueprint goal**, and **66,666 language destination goal** are capacities or ambitions, not delivered counts. The ONE app, Factory, TABOR WORLD, TaborTales, SSS-CS and other distinct products retain independent source and lifecycle rules.

## Original-source boundaries

- Every original user-authored source file remains unchanged by this atlas.
- No prior app code was copied into, rewritten by, or replaced with the recovery atlas.
- Historically named systems without accessible public original source are shown as references or owner-held anchors, not implemented code.
- Exact source quotations belong in owner-controlled originals and may be viewed only through explicit local file selection; analyst-authored labels and descriptions are separate derived metadata.
- Historical/theological/technical assertions remain author-attributed or unverified until sourced.
- No payment, medical-provider connectivity, cloud service, real-time media generation, or app store release is implied.

## Development validation

Run: node --test app-factory/nathan-ai/build-recovery/atlas.test.cjs

All public records are rendered via DOM textContent rather than interpreted as arbitrary HTML. Original locally selected files remain browser-private and are never uploaded. Browser/mobile testing and GitHub Pages rollout are separate from successful repository/CI checks.

**Motto:** Preserve → Index → Relate → Derive; Never Replace. 1+1=2 BUILD. 1-1=0 CLEAR.
