mcp_email_reader

@karateboss

About mcp_email_reader

Model Context Protocol (MCP) server to connect to an email server and to read emails

Config

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

{
  "mcpServers": {
    "mcp_email_reader": {
      "command": "uvx",
      "args": [
        "--from",
        "git+https://github.com/karateboss/mcp_email_reader@main",
        "mcp_email_reader"
      ],
      "env": {
        "IMAP_SERVER": "",
        "EMAIL_ACCOUNT": "",
        "EMAIL_PASSWORD_ENC": "",
        "EMAIL_SECRET_KEY": ""
      }
    }
  }
}

Tools

No tools detected

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

Overview

What is mcp_email_reader?

mcp_email_reader is a Model Context Protocol (MCP) server that exposes email-reading tools—search_emails, download_attachment, and list_folders—for use with MCP-enabled AI tools like Claude Desktop or LibreChat with Ollama.

How to use mcp_email_reader?

Install Python 3.10+, install uv, then add a JSON configuration block to your claude_desktop_config.json specifying the uvx command, the repository URL, and environment variables for your IMAP server, email account, encrypted password, and secret key.

Key features of mcp_email_reader

Search emails using IMAP via search_emails tool
Download email attachments with download_attachment
List email folders with list_folders
Works with Claude Desktop and LibreChat (Ollama)
Uses encrypted password and secret key for authentication

Use cases of mcp_email_reader

Let an AI assistant search through your inbox for specific emails
Automatically download attachments from emails using natural language commands
List all email folders to organize or find messages across accounts

FAQ from mcp_email_reader

What tools does mcp_email_reader provide?

It provides three MCP tools: search_emails, download_attachment, and list_folders.

What are the runtime requirements?

Python 3.10 or higher and the uv package manager are required. The server is tested on Windows and macOS.

How is my email password secured?

The password is stored encrypted using a secret key; both the encrypted password and secret key are placed in the local claude_desktop_config.json file. The README notes this is a potential security risk.

What email protocols does it support?

It uses IMAP for fetching emails. You must provide your IMAP server address.

Can I use it without Claude Desktop?

Yes, it works with any MCP-enabled AI tool; the README specifically mentions LibreChat with Ollama.

Frequently asked questions

What tools does mcp_email_reader provide?

It provides three MCP tools: `search_emails`, `download_attachment`, and `list_folders`.

What are the runtime requirements?

Python 3.10 or higher and the `uv` package manager are required. The server is tested on Windows and macOS.

How is my email password secured?

The password is stored encrypted using a secret key; both the encrypted password and secret key are placed in the local `claude_desktop_config.json` file. The README notes this is a potential security risk.

What email protocols does it support?

It uses IMAP for fetching emails. You must provide your IMAP server address.

Can I use it without Claude Desktop?

Yes, it works with any MCP-enabled AI tool; the README specifically mentions LibreChat with Ollama.

Comments

Basic information

More Communication MCP servers

Caulo

Patry

Caulo is the social layer of the agentic web: a social network your AI agent reads, briefs you on, and posts to — by prompt. Ad-free, verified humans only, with provenance on every post (human / agent-assisted / agent-au

Socialync

Jack Vitick

Draft, schedule, and publish social media posts to 8 platforms from any AI agent: Facebook, Instagram, X, TikTok, YouTube, LinkedIn, Threads, and Bluesky. 19 tools covering drafting, scheduling, publishing, media upload,

Mcp Telegram

dryeab

MCP Server for Telegram

244

Outlook MCP Server

Wallisking1991

AgentCouch

Messaging rooms for AI agents: hand off context across tools, worktrees, machines, and teammates. Agents meet in shared rooms over MCP, hand off directly, and leave a replayable transcript; live sessions wake on reply, b