Bookmark Manager MCP Server

A lightweight MCP (Model Context Protocol) server that lets you save, tag, and search bookmarks directly from Claude Code. Instead of switching to a browser or separate app, you can manage your bookmarks conversationally — just tell Claude to save a link or find one you saved earlier.

Why It's Useful

• Stay in your workflow — save and retrieve bookmarks without leaving the terminal

• Natural language search — search by keyword across URLs, titles, and tags

• Zero setup storage — bookmarks persist in a local JSON file, no database required

• Simple and extensible — minimal codebase that's easy to customize

Related MCP server: Anybox MCP Server

Installation

Prerequisites

• Node.js v18 or higher

• Claude Code CLI installed

Steps

1. Clone or download this project:

   cd ~/mcp-projects
   git clone <your-repo-url> my-first-mcp
   cd my-first-mcp

2. Install dependencies:

   npm install

3. Add the server to your Claude Code MCP configuration. In your project's .mcp.json:

   {
     "mcpServers": {
       "bookmark-manager": {
         "command": "node",
         "args": ["index.js"],
         "cwd": "/path/to/my-first-mcp"
       }
     }
   }

4. Start Claude Code in your project directory — the server will be available automatically.

Usage Examples

Save a bookmark with a title

"Add a bookmark to https://react.dev with the title React Documentation"

Bookmark saved!
  ID:    1
  URL:   https://react.dev
  Title: React Documentation
  Tags:  (none)

Save a bookmark with tags

"Bookmark https://expressjs.com, title Express.js, tags: backend,nodejs,framework"

Bookmark saved!
  ID:    2
  URL:   https://expressjs.com
  Title: Express.js
  Tags:  backend, nodejs, framework

Search your bookmarks

"Search my bookmarks for react"

Found 1 bookmark(s):
• [1] React Documentation
  https://react.dev
  Tags: (none)

Limitations and Known Issues

• No delete or edit — bookmarks can be saved and searched, but there is currently no tool to update or remove individual bookmarks. You would need to manually edit bookmarks.json.

• File-based storage — all bookmarks are stored in a single bookmarks.json file. This works well for personal use but is not designed for large-scale or multi-user scenarios.

• Search is basic — search performs case-insensitive substring matching across URL, title, and tags. There is no fuzzy matching or ranking.

• No duplicate detection — saving the same URL twice will create two separate entries.