Moist
Enables Gmail operations such as reading, sending, searching, and managing emails, threads, labels, and drafts via MCP tools.
README
Moist
Gmail MCP Connector
"Neither rain nor snow nor stranded tokens shall stay this messenger."
Moist is an MCP server that exposes Gmail operations as tools. It handles OAuth 2.0, token refresh, rate limiting, and MIME parsing so consuming applications don't have to.
Quick Start
1. Google Cloud Setup
- Create a project in the Google Cloud Console
- Enable the Gmail API
- Create OAuth 2.0 Client ID credentials (Desktop application type)
- Add
http://localhost:3000/oauth/callbackas an authorized redirect URI - Note your Client ID and Client Secret
2. MCP Configuration
Add Moist to your MCP client configuration:
{
"mcpServers": {
"moist": {
"command": "npx",
"args": ["@ticktockbent/moist"],
"env": {
"MOIST_CLIENT_ID": "your_client_id",
"MOIST_CLIENT_SECRET": "your_client_secret"
}
}
}
}
3. First Run
On the first connection, Moist opens your browser for Google OAuth consent. Once authorized, tokens are encrypted and stored locally at ~/.moist/tokens.json. Subsequent runs authenticate automatically.
Environment Variables
| Variable | Required | Description |
|---|---|---|
MOIST_CLIENT_ID |
Yes | Google OAuth 2.0 Client ID |
MOIST_CLIENT_SECRET |
Yes | Google OAuth 2.0 Client Secret |
MOIST_REDIRECT_URI |
No | OAuth callback URI (default: http://localhost:3000/oauth/callback) |
Tools
Authentication
| Tool | Description |
|---|---|
moist_auth_status |
Check authentication state, email, scopes, and token expiry |
moist_auth_logout |
Revoke tokens and clear stored credentials |
Messages
| Tool | Description |
|---|---|
moist_list_messages |
Search and list messages with pagination and label filtering |
moist_get_message |
Get full message details including body, headers, and attachments |
moist_send_message |
Send an email (supports replies via replyTo and threading via threadId) |
moist_trash_message |
Move a message to trash |
moist_delete_message |
Permanently delete a message (irreversible) |
Threads
| Tool | Description |
|---|---|
moist_list_threads |
List threads with search and label filtering |
moist_get_thread |
Get a thread with all its messages |
moist_trash_thread |
Move an entire thread to trash |
Labels
| Tool | Description |
|---|---|
moist_list_labels |
List all labels (system and user-created) |
moist_modify_labels |
Add or remove labels from a message |
Search
| Tool | Description |
|---|---|
moist_search |
Search messages using Gmail's full query syntax |
Gmail search supports operators like from:, to:, subject:, has:attachment, is:unread, is:starred, label:, after:, before:, filename:, and "exact phrases". Operators can be combined:
from:bank@example.com has:attachment after:2024/01/01 subject:statement
Drafts
| Tool | Description |
|---|---|
moist_create_draft |
Create a new draft email |
moist_list_drafts |
List drafts with pagination |
moist_delete_draft |
Delete a draft |
moist_send_draft |
Send an existing draft |
Architecture
src/
├── index.ts # MCP server entry point (stdio transport)
├── types.ts # TypeScript interfaces
├── auth/
│ ├── oauth.ts # OAuth 2.0 flow, token refresh, browser consent
│ └── storage.ts # Encrypted token persistence (~/.moist/)
├── client/
│ ├── gmail.ts # Gmail API wrapper, MIME parsing, message building
│ └── rate-limiter.ts # Sliding-window quota tracker
└── tools/
├── auth.ts # Auth status and logout tools
├── messages.ts # Message CRUD tools
├── threads.ts # Thread tools
├── labels.ts # Label tools
├── search.ts # Search tool
└── drafts.ts # Draft tools
Key Design Decisions
- Stateless -- No caching. Each tool call hits the Gmail API directly.
- Single account -- One Gmail account per server instance. Run multiple instances for multiple accounts.
- Raw content -- Message bodies are returned as-is (plain text and HTML). No parsing or summarization.
- Encrypted storage -- Tokens at rest are AES-256-CBC encrypted using a machine-specific derived key.
- Rate limiting -- Tracks Gmail API quota usage in a sliding 1-second window (250 units/sec per Google's limits). Returns structured errors with retry timing when exceeded.
Error Handling
All tools return consistent error shapes:
{
"error": "not_found",
"message": "Human-readable description",
"details": {}
}
| Error Code | Meaning |
|---|---|
not_found |
Message, thread, or label doesn't exist |
rate_limited |
Gmail API quota exceeded (includes retryAfter) |
auth_failed |
Token expired or revoked |
invalid_request |
Bad parameters |
api_error |
Gmail API error (details included) |
OAuth Scopes
Moist requests these scopes during authorization:
| Scope | Purpose |
|---|---|
gmail.readonly |
Read messages, threads, and labels |
gmail.send |
Send messages |
gmail.modify |
Modify labels, trash/untrash |
gmail.compose |
Create and manage drafts |
Development
# Install dependencies
npm install
# Build
npm run build
# Watch mode
npm run dev
License
MIT
References
- Gmail API Documentation
- MCP Specification
- Going Postal -- The source of the name
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.