WikiStub-Seed

EN | DE | ES | JA | RU | ZH

WikiStub-Seed is a multilingual-ready JSON knowledge framework for AI-assisted research, documentation, learning systems and LLM workflows. It ships a curated wikistub_seed.json dataset with 630 compact German/English knowledge stubs across 12 scientific and cultural domains, prepared ES/ZH/JA/RU language slots, plus Python tools for validation, export and translation.

WikiStub-Seed is a knowledge-stub seed library, not a wiki.

What It Contains

• 630 knowledge stubs in wikistub_seed.json with DE/EN content and prepared ES/ZH/JA/RU language slots
• 12 top-level domains, including mathematics, physics, chemistry, biology, medicine, psychology, AI, engineering, society, economics, history and culture
• 85 subcategories with short, neutral definitions and relevance notes
• Canonical definitions.{lang} and relevance_i18n.{lang} maps while retaining legacy definition_de, definition_en and relevance
• Python CLI tooling for statistics, validation, consistency checks and Markdown export
• A documented wikistub-seed-data-v1 export direction for future static Web/PWA use
• No required external dependencies for core import, export, validation or CLI use

Use Cases

• Seed a local knowledge base for AI-assisted writing or research
• Build documentation glossaries, learning maps or concept catalogs
• Export structured Markdown for Obsidian, GitHub Pages or static sites
• Feed retrieval, embeddings or LLM context pipelines with compact domain stubs
• Translate and extend a domain-neutral knowledge skeleton in a controlled JSON format

Data Shape

Each stub is intentionally small and machine-readable:

{
  "title": "Domain-Driven Design",
  "definition_de": "Ein Ansatz zur Modellierung komplexer Software, der die Fachdomäne in den Mittelpunkt stellt.",
  "definition_en": "An approach to modeling complex software that places the business domain at the center of development.",
  "relevance": "Hilft, komplexe Systeme verständlich und wartbar zu gestalten.",
  "definitions": {
    "de": "Ein Ansatz zur Modellierung komplexer Software, der die Fachdomäne in den Mittelpunkt stellt.",
    "en": "An approach to modeling complex software that places the business domain at the center of development.",
    "es": "",
    "zh": "",
    "ja": "",
    "ru": ""
  },
  "relevance_i18n": {
    "de": "Hilft, komplexe Systeme verständlich und wartbar zu gestalten.",
    "en": "",
    "es": "",
    "zh": "",
    "ja": "",
    "ru": ""
  },
  "tags": ["Informatik", "Software Engineering"]
}

The current authoritative source is wikistub_seed.json. EXPORTFORMAT.md documents the stable wrapper format wikistub-seed-data-v1 for Web/PWA, API and LLM exports.

Quick Start

git clone https://github.com/dev-bricks/WikiStub-Seed.git
cd WikiStub-Seed

python wikistub_seed_cli.py --help
python wikistub_seed_cli.py stats
python wikistub_seed_cli.py check
python wikistub_seed_pipeline.py validate
python wikistub_seed_pipeline.py export --output --english

On Windows, start.bat opens the CLI entry point. Exported files are written to output/; that folder is local and not versioned.

Core Commands

Command	Purpose
python wikistub_seed_cli.py stats	Print stub, category and tag statistics
python wikistub_seed_cli.py check	Run consistency checks over the JSON dataset
python wikistub_seed_pipeline.py validate	Validate the pipeline input data
python wikistub_seed_pipeline.py export --output --english	Export the JSON dataset to Markdown
python wikistub_seed_pipeline.py translate	Optionally translate missing English definitions when configured

Repository Map

Path	Purpose
wikistub_seed.json	Authoritative bilingual knowledge dataset
01_Mathematik/ ... 12_Kultur_Kunst_Sprache/	Domain-oriented Markdown source/export structure
wikistub_seed_cli.py	CLI for stats and checks
wikistub_seed_pipeline.py	Import, export, validation and optional translation pipeline
md_to_json.py	Markdown-to-JSON import helper
check_duplicates.py	Duplicate/consistency helper
EXPORTFORMAT.md	Stable exchange-format plan
web_publisher/	Static Web/PWA publisher (PWA, offline cache, search, DE/EN toggle)

Privacy

WikiStub-Seed is local-first. Core usage reads and writes local JSON/Markdown files only. There is no telemetry and no automatic network communication.

The optional translation command can call an external API only when ANTHROPIC_API_KEY is set and the optional anthropic package is installed.

Roadmap

Completed:

• 12 top-level domains and 85 subcategories
• 630 bilingual stubs in a single JSON master file
• Markdown export and JSON synchronization tooling
• CLI smoke tests in GitHub Actions, plus dedicated macOS/Linux source smokes for wikistub_seed_cli.py check and wikistub_seed_pipeline.py validate
• Static Web/PWA publisher with search and offline cache (web_publisher/)
• wikistub-seed-data-v1 schema wrapper with prepared DE/EN/ES/ZH/JA/RU language slots

Planned:

• Unified tag cleanup
• Obsidian/GitHub Pages export paths
• Optional embeddings and search API

Deutsch

WikiStub-Seed ist ein mehrsprachig vorbereitetes JSON-Wissensgerüst für KI-gestützte Wissensarbeit. Das Repository enthält 630 kompakte Wissens-Stubs mit Deutsch/Englisch-Inhalten und vorbereiteten Sprachslots für Spanisch, Chinesisch, Japanisch und Russisch, verteilt auf 12 Wissenschafts- und Kulturbereiche. Die Stubs sind kurz, neutral, versionierbar und für Automatisierung, Dokumentation, Lernsysteme und LLM-Kontexte geeignet.

WikiStub-Seed arbeitet standardmäßig lokal mit wikistub_seed.json. Die Kernfunktionen benötigen keine externen Pakete. Nur die optionale Übersetzungsfunktion nutzt externe API-Aufrufe, wenn ein API-Key gesetzt und das optionale Paket installiert wurde.

Wichtige Einstiegspunkte:

• python wikistub_seed_cli.py stats zeigt Statistik und Kategorien.
• python wikistub_seed_cli.py check prüft den Datenbestand.
• python wikistub_seed_pipeline.py export --output --english exportiert Markdown.
• EXPORTFORMAT.md beschreibt den geplanten stabilen Austauschstandard.
• web_publisher/ enthält den fertigen statischen Web/PWA-Publisher mit Offline-Cache und DE/EN-Toggle.

License

MIT License. See LICENSE.

This project is an unpaid open-source donation. Liability is limited to intent and gross negligence under Section 521 of the German Civil Code, with the MIT License disclaimers applying as well. Use at your own risk.