Memento

Name: Memento
Author: iAchilles

@iAchilles

Repository

About Memento

A local, fully-offline MCP memory server using SQLite + FTS5 + sqlite-vec with embedding support

#mcp #server #memory #sqlite #fts5 #embedding

Config

Add this server to your MCP-compatible client using the configuration below.

{
  "mcpServers": {
    "memory": {
      "description": "Custom memory backed by SQLite + vec + FTS5 with embedding support",
      "command": "npx",
      "args": [
        "-y",
        "memento"
      ],
      "env": {
        "MEMORY_DB_PATH": "/Path/To/Your/memory.db"
      },
      "options": {
        "autoStart": true,
        "restartOnCrash": true
      }
    }
  }
}

Tools

No tools detected

Fetch the live tool list by running this server in a temporary sandbox using the button above.

Overview

What is Memento?

Memento is a local, fully-offline MCP memory server that uses SQLite with FTS5 and sqlite-vec for keyword and semantic vector search. Built for AI agents, it stores a structured graph of entities, observations, and relations, and uses an offline embedding model (bge-m3 via @xenova/transformers). It is intended for developers who need persistent, private memory in their MCP-based workflows.

How to use Memento?

Install globally: npm install -g @iachilles/memento. Set the MEMORY_DB_PATH environment variable to your desired database file path, then run memento from the terminal. For Claude Desktop integration, add a tool entry with command npx -y memento and the MEMORY_DB_PATH in the environment block

Comments

Basic information

More Memory & Knowledge MCP servers

Obsidian MCP Server

cyanheads

Read, write, search, and surgically edit Obsidian vault notes, tags, and frontmatter via MCP. STDIO or Streamable HTTP.

618

Solomd

zhitongblog

A markdown editor — and the bridge to your LLM. Local-first, MIT, ~15 MB. Bundled MCP server lets Claude Code / Codex / Cursor drive your vault directly. 14 AI providers BYOK.

425