MoonshotAI
Featured
Kimi-K3
Kimi K3 is an open-weight, 2.8T-parameter native multimodal agentic model with a 1M-token context window, designed for frontier coding, knowledge work, and reasoning tasks.
AI repository intelligence
FR-AI / ANALYSISWhy it is worth attention
It extends Karpathy's LLM Wiki pattern with explicit provenance roles, a published-product layer, a staging area for AI output, and typed machine-readable links, all while keeping Obsidian and Codex as optional interfaces and the repository itself as the durable system.
Who it is for
- Researchers who need to keep personal ideas distinct from external evidence
- Writers and publishers who create AI-assisted content with citation management
- Teachers developing course materials with a review pipeline
- AI power users who want a structured knowledge base that works with Codex
Use cases
- Capturing and indexing personal notes and external sources with provenance
- Composing articles or courses from a knowledge base with source planning and citation
- Migrating an existing Obsidian vault into a review-first, agent-friendly structure
- Querying the knowledge graph to find original sources and relationships
Strengths
- Clear separation of raw input, external sources, published outputs, and wiki concepts via dedicated directories
- Staged review lifecycle prevents AI-generated text from becoming truth without human approval
- Typed YAML sidecars in .kb/links/ provide a machine-readable graph that can be queried independently
- Works with plain Markdown and optional Obsidian or Codex, making it accessible without vendor lock-in
Considerations
- Automation is intentionally minimal; no built-in search, embeddings, or graph database until real queries justify them
- Requires consistent human review for every promotion from staging to wiki or products
- Obsidian and Codex are optional but the workflows are designed around them; using only plain Markdown loses some visual and agent integration
README quick start
AI Content Knowledge Base A review-first, AI-native knowledge base for people who research, write, teach, and publish. Markdown + Codex + Obsidian + typed YAML relationships
English · 简体中文
[!NOTE] This project was inspired by Andrej Karpathy's LLM Wiki pattern: immutable raw sources, a persistent and compounding Markdown wiki maintained by an LLM, and a repository schema such as
AGENTS.mdthat defines how the agent should ingest, query, and maintain knowledge.
ai-content-kbextends that idea with explicit provenance roles, a first-class published-product layer, review-first staging, typed machine-readable relationships, content-creation workflows, and safe legacy-vault migration.
Most knowledge bases mix personal thinking, external evidence, published work, and AI-generated text in the same pile. This template gives each a clear role—and gives AI agents rules for working without silently polluting originals or publication workflows.
Originals stay trustworthy. The wiki stays readable. The graph stays queryable. AI output stays reviewable.
The screenshot shows a mature vault using this architecture. A fresh clone starts with a small synthetic example and grows as you add reviewed notes, links, and tags.
Why this project
| Need | How this project handles it |
|---|---|
| Keep personal ideas distinct from external claims | Separate raw/ and sources/ provenance layers |
| Reuse published articles, courses, and scripts | Treat reviewed output as a first-class products/ layer |
| Let people browse without turning summaries into proof | Use wiki/ as a cited human interface |
| Give agents precise relationships | Store typed, evidenced YAML sidecars in .kb/links/ |
| Prevent generated text from becoming truth by accident | Route AI output through .kb/staging/ and human review |
| Work naturally with Codex | Ship durable workflows in root AGENTS.md |
| Keep Obsidian optional | Store everything as ordinary Markdown, YAML, JSON, and folders |
Quickstart
git clone https://github.com/mrbear1024/ai-content-kb.git
cd ai-content-kb
No application is required to read the vault. Use plain Markdown, open the root as an Obsidian vault, or use the intended Codex workflow below.
Open in Code
Description
A review-first reference architecture for AI-assisted personal content knowledge systems
Related repositories
Similar projects matched by category, topics, and programming language.
xuchonglang
Featured
investing-for-beginners
A structured investing guide for Chinese beginners covering US stocks, options, and cryptocurrency, with focus on foundational concepts and risk awareness.
Krishnagangwal
Featured
CS-Fundamentals
A curated collection of Computer Science fundamentals (PDFs, notes, cheatsheets, interview question banks) for placement preparation, covering seven core subjects plus general resources.