Obsidian LLM Wiki FAQs

Question 1

What is 'Karpathy-style Knowledge Compilation' and how does it work?

Accepted Answer

Inspired by Andrej Karpathy, this feature automates the process of extracting, structuring, and interlinking information from raw sources (like papers or articles) into a coherent, cross-referenced wiki within your Obsidian vault, handled by an LLM and helper scripts.

Question 2

Is my Obsidian vault safe from unintended changes by AI agents?

Accepted Answer

Yes, security is paramount. All write operations initiated by AI agents are dry-run by default, requiring explicit confirmation (`dryRun: false`) to make changes. Additionally, path traversal is blocked, and critical `.obsidian/` directories are protected from writes.

Question 3

How does Obsidian LLM Wiki integrate with AI agents?

Accepted Answer

It transforms your Obsidian vault into an MCP server, allowing compatible AI agents like Claude Code, Cursor, and Windsurf to seamlessly read, search, create, and generate content directly from your notes.

Question 4

Can AI agents access my Obsidian notes when Obsidian is closed?

Accepted Answer

Yes, Obsidian LLM Wiki features a unique filesystem fallback mechanism that enables AI agents to interact with your vault even when the Obsidian application is not running, ensuring continuous access and functionality.

Question 5

What unique advantages does Obsidian LLM Wiki offer over other Obsidian AI integrations?

Accepted Answer

It provides a comprehensive suite of features including filesystem fallback (offline operation), advanced knowledge graph queries, built-in Karpathy-style compilation, vault health checks, batch operations, real-time events, and dry-run safety for all write actions.

Obsidian LLM Wiki

Obsidian LLM Wiki

Key Features

Use Cases

Key Features

Use Cases