mcpField
E
email-tool

email-tool

Local
github.com/skhan61Communication0Updated 1d ago
Repository

About email-tool

email-tool is an MCP server for AI assistants that need to manage mail through IMAP and SMTP. It runs locally, stores passwords in the operating system keyring, and includes audit logging, approval before sending, and domain restrictions for CUI-marked mail.

Key features

  1. 01Read, search, draft, and send email
  2. 02IMAP and SMTP support
  3. 03Handle attachments
  4. 04Store passwords in the OS keyring
  5. 05Audit log and sending approval
  6. 06Domain restrictions for CUI-marked mail

Use cases

  1. 01Searching and reading a mailbox with an AI assistant
  2. 02Preparing messages for approval before sending
  3. 03Managing email under CMMC Level 2 and NIST SP 800-171 safeguards

MCP support

It exposes email actions to MCP clients, including reading, searching, drafting, and sending messages over IMAP/SMTP.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

mcp.json · Universal config
{
  "mcpServers": {
    "skhan61-email-tool": {
      "command": "uvx",
      "args": [
        "email-tool-mcp"
      ]
    }
  }
}
Run commanduvx email-tool-mcp

Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.

Packages
pypiemail-tool-mcp5.6.0

Tools

15

Saved tool list · updated today

This tool takes no parameters.

Details

From the project README

email-tool icon

email-tool

Email for any AI agent that speaks MCP: Claude Code, Claude Desktop, Cowork, ChatGPT desktop, Codex and others.
Read, search, draft and send mail, with attachments, over IMAP/SMTP.

Works with
Claude Code plugin Claude Desktop and Cowork ChatGPT desktop and Codex MCP server

Compliance
CMMC Level 2 / NIST SP 800-171 control mapping

Build and security
CI CodeQL OpenSSF Scorecard Latest release

Code quality
Python version Tested on Linux, macOS and Windows Ruff mypy strict uv pre-commit Dependabot

Project
License Security policy Contributing

email-tool lets an AI assistant read, search, draft and send your email. It runs on your own computer, keeps passwords in your operating system's keyring, and is built for organizations under CMMC Level 2 and NIST SP 800-171: every action is logged, sending can wait for your approval, and CUI-marked mail stays inside the domains you allow.

Documentation: skhan61.github.io/email-tool

Quick start

No terminal, any app: Start here. Install email-tool in your app (Claude Desktop and Cowork: one file to open; ChatGPT desktop and Codex: one installer to open), then ask your assistant to "add my mailbox".

From a terminal:

  1. Install uv.

  2. Connect your AI app. In Claude Code:

    /plugin marketplace add skhan61/email-tool
/plugin install email-tool@email-tool

    Other apps: Advanced setup.

  3. Ask your assistant to "add my mailbox", or from a terminal:

    uv tool install email-tool-mcp
email-tool-mcp add work --auth microsoft --address [email protected]

    Gmail, Google Workspace and other accounts: Mailbox details.

Documentation

PageWhat it covers
Start hereInstall in any app and add a mailbox, no terminal
Everyday useWhat to ask, the tools, approving before sending
Mailbox detailsWhat each mailbox needs (Gmail App Password, Microsoft, GCC High, Google Workspace, others), and the terminal steps
Advanced setupuv by hand, configuration files, ChatGPT Business, other MCP apps, update or remove
Why CMMCEach protection, with examples
CMMC control mappingNIST SP 800-171 requirements, evidence and your part
For administratorsPolicy, audit log, FIPS mode, your own Microsoft or Google app
TroubleshootingEvery message and its fix
PrivacyWhat email-tool accesses and where it goes

Contributing and security

See CONTRIBUTING.md. Report vulnerabilities privately as described in SECURITY.md.

License

MIT

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.