Question 1

What types of content can Memex index?

Accepted Answer

Memex indexes markdown vaults, processing text, wikilinks, and frontmatter (aliases, tags) for comprehensive search. It automatically re-indexes changed files to ensure your knowledge base is always up-to-date, excluding hidden directories.

Question 2

What is Memex?

Accepted Answer

Memex equips large language models (LLMs) with persistent memory and advanced knowledge retrieval capabilities. It indexes your markdown vaults for semantic and full-text search, enabling LLMs to learn, document findings, and recall information across sessions.

Question 3

Is Memex compatible with Obsidian vaults?

Accepted Answer

Yes, Memex is designed to work seamlessly with markdown vaults, including those managed by Obsidian. It indexes your notes and wikilinks, enhancing your existing 'second brain' by making its content accessible and searchable for LLMs.

Question 4

How does Memex integrate with LLMs, especially Claude Code?

Accepted Answer

Memex integrates as an MCP server for Claude Code, providing direct tools like `search()` for semantic queries and `explore()` for graph traversal. This allows Claude to seamlessly interact with and leverage your knowledge base.

Question 5

How does Memex provide persistent memory to LLMs?

Accepted Answer

It creates a local index of your markdown vaults, including full-text search, embeddings for semantic similarity, and a wikilink graph. LLMs can then query this index using natural language tools to retrieve relevant information and build context.

Memex

Client Configuration

System Overview

Frequently Asked Questions

Related MCP Servers

Carbone

CloudCDN

Cloudflare AI Agent Server