mcp-email-server
IMAP and SMTP via MCP Server
Documentation
mcp-email-server
An MCP server for reading, searching, organizing, and sending email through
IMAP and SMTP.
> [!NOTE]
> Version 1.0.0 introduces Local Email App V2. Updating the package does not
> automatically import existing settings created with PyPI 0.16.0 and earlier:
> they remain active in backward-compatible `legacy` mode, so there is no required
> migration. If you
> would like to use the new managed storage, you can review and import those
> settings whenever it is convenient.
`mcp-email-server` supports Windows, macOS, and Linux. See
Security for platform-specific filesystem and credential
storage details.
Optional migration for existing installations
An `@latest` release that includes Local Email App V2 offers a preview-first
CLI migration:
uvx mcp-email-server@latest config init \
--database ~/.config/mcp-email-server/managed.sqlite3
uvx mcp-email-server@latest config import-legacy
uvx mcp-email-server@latest config import-legacy --apply
uvx mcp-email-server@latest config doctorThe apply step displays the plan again and asks for `IMPORT` confirmation. A
complete import selects managed mode; otherwise the existing legacy settings
remain selected. The source TOML file and its legacy keyring entries are left
untouched. You can also run `uvx mcp-email-server@latest ui` and choose **Import
existing settings**. After a successful import, restart running MCP clients. See
the detailed upgrade guidance
Quick start
1. Configure an email account
From this source checkout, run the configuration UI with
`uv`:
uv sync
uv run mcp-email-server uiFor a published release whose notes state that it includes Local Email App V2,
`uvx mcp-email-server@latest ui` is the equivalent temporary invocation.
Keep the foreground command running. On a truly empty installation, the
authenticated browser session prepares private account storage at the safe local
default; existing TOML or environment configuration instead offers an explicit
import review while the previous settings keep running. The account-first UI has
only Email accounts and Settings & help as primary destinations. Start
with the email address and password; the UI fills common connection settings from
the email domain and keeps them editable, while outgoing mail remains optional.
A saved complete account is ready without a separate activation step. Use
Password & test on the saved account if desired, then restart the MCP client
to apply the selected settings.
2. Configure the MCP client
Use the same V2-capable distribution for stdio as for the UI. For a published
V2 release, add the following server definition to the MCP client:
{
"mcpServers": {
"mcp-email-server": {
"command": "uvx",
"args": ["mcp-email-server@latest", "stdio"]
}
}
}Restart the MCP client after updating its configuration. When testing this
source checkout before publication, invoke `uv run --directory
/absolute/path/to/mcp-email-server mcp-email-server stdio` instead of pairing a
managed catalog with PyPI `@latest`.
3. Verify the connection
Ask the client to list the configured email accounts or recent messages.
Other configuration methods
For the SQLite-backed managed CLI workflow, Windows and POSIX storage boundaries,
headless environments, containers, multiple accounts, custom TLS settings, and
environment-variable configuration, see the
Documentation
Development
See CONTRIBUTING.md.
License
This project is licensed under the terms of the LICENSE.
Frequently asked questions
What is mcp-email-server?
mcp-email-server is IMAP and SMTP via MCP Server
How do I install mcp-email-server?
Open the GitHub repository and follow its README. Most MCP servers are added to your client's MCP config, then called by your agent.
Is mcp-email-server open source?
Yes — it is hosted on GitHub at https://github.com/ai-zerolab/mcp-email-server and has 96 stars.
Related MCP tools
[](https://github.com/chigwell/telegram-mcp/actions/workflows/python-lint-format.yml)
🙌 OpenHands: Code Less, Make More for the Model Context Protocol. Enhance AI assistants with powerful integrations. Python-based implementation.
Universal memory layer for AI Agents; Announcing OpenMemory MCP - local and secure memory management. Python-based implementation.
基于大模型搭建的聊天机器人,同时支持 微信公众号、企业微信应用、飞书、钉钉 等接入,可选择ChatGPT/Claude/DeepSeek/文心一言/讯飞星火/通义千问/ Gemini/GLM-4/Kimi/LinkAI,能处理文本、语音和图片,访问操作系统和互联网,支持基于自有知识库进行定制企业智能客服。
An LLM agent that conducts deep research (local and web) on any given topic and generates a long report with citations. Built for the Model Context Protocol to
🚀 The fast, Pythonic way to build MCP servers and clients Trusted by 19900+ developers. Trusted by 19900+ developers. Trusted by 19900+ developers.
Run your own MCP server? See who uses it and what to fix.
Measure it with TrackMCP