mcpField
D
DragApp — Gmail & Whatsapp For AI Tools (Shared Inbox MCP Server)

DragApp — Gmail & Whatsapp For AI Tools (Shared Inbox MCP Server)

Local
github.com/nick-timmsCommunication1Updated 4w ago

About DragApp — Gmail & Whatsapp For AI Tools (Shared Inbox MCP Server)

DragApp MCP Server lets teams manage DragApp shared email and WhatsApp inboxes through MCP-compatible AI clients. It covers conversation handling alongside boards, knowledge-base content, tasks, and analytics.

Key features

  1. 01Read, search, and reply to email threads
  2. 02Send WhatsApp messages and view conversations
  3. 03Manage boards, cards, labels, and contacts
  4. 04Create and update knowledge-base articles
  5. 05Review response-time and activity analytics
  6. 06Manage automations and tasks

Use cases

  1. 01Find unread messages on a support board
  2. 02Draft replies to customer emails
  3. 03Move or assign conversations and tasks
  4. 04Search help articles and team response metrics

MCP support

Exposes DragApp email, WhatsApp, boards, cards, knowledge base, analytics, and other operations as tools for MCP clients. It supports a hosted remote connector and a locally configured server.

Configuration

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

mcp.json · Universal config
{
"mcpServers": {
"drag": {
"command": "npx",
"args": [
"-y",
"@dragapp/mcp-server"
],
"env": {
"DRAG_API_KEY": "<DRAG_API_KEY>"
}
}
}
}
Run commandnpx -y @dragapp/mcp-server
Environment variables to set
DRAG_API_KEYSecretYour DragApp API key. Get it from app.dragapp.com → Settings → Integrations.

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
npm@dragapp/mcp-server0.1.5

Tools

47

Saved tool list · updated 3w ago

This tool takes no parameters.

Details

From the project README

DragApp MCP Server

An MCP server for DragApp — AI-powered email & WhatsApp for teams. Read and reply to emails, send WhatsApp messages, search threads, manage boards, and more from Claude, ChatGPT, Cursor, or any MCP-compatible AI tool.

Setup

Fastest: connect from your AI client (hosted)

Add this URL as a custom connector:

https://app.dragapp.com/mcp
  • Claude — Settings → Connectors → Add custom connector → paste the URL
  • ChatGPT — Settings → Connectors → Add → paste the URL
  • Gemini — add it as a remote MCP server with the same URL

When you connect, a Drag page opens and asks for your API key once (DragApp → Settings → Integrations). The key is verified and handed to your client as its access token; the service stores nothing, and the connect flow never puts your key in a URL. Works in any client that supports remote MCP connectors (OAuth 2.1 + PKCE, dynamic client registration). Nothing to install.

Local (config-file clients: Cursor, Windsurf, Claude Code)
1. Get your API key

Go to DragApp → Settings → Integrations → copy your API key.

2. Connect to your AI tool

Claude Desktop — add to ~/Library/Application Support/Claude/claude_desktop_config.json:

{
  "mcpServers": {
    "dragapp": {
      "command": "npx",
      "args": ["-y", "@dragapp/mcp-server"],
      "env": {
        "DRAG_API_KEY": "your-api-key"
      }
    }
  }
}

Cursor — add to .cursor/mcp.json:

{
  "mcpServers": {
    "dragapp": {
      "command": "npx",
      "args": ["-y", "@dragapp/mcp-server"],
      "env": {
        "DRAG_API_KEY": "your-api-key"
      }
    }
  }
}

Works with any MCP-compatible client — ChatGPT, Windsurf, Claude Code, etc.

What you can do

Ask your AI assistant:

  • "Show me unread emails on the Support board"
  • "Read the latest email from Acme Corp and draft a reply"
  • "Move all billing threads to the Done column"
  • "Search for emails mentioning 'invoice' on the Sales board"
  • "What's my team's average response time this week?"
  • "List all knowledge base articles"
  • "Search our help center for how invoices work"
  • "Assign the newest email on the Support board to Sarah"
  • "Create a task on the Support board assigned to Sarah"

Tools

Email (8 tools)

list_threads · get_thread · reply_to_thread · send_new_email · search_threads · filter_threads · move_thread · move_threads_bulk

Boards (5 tools)

list_boards · get_board · list_columns · list_board_members · list_teams

Cards (6 tools)

list_cards_in_column · get_card · create_card · update_card · move_card · archive_card

Labels (4 tools)

list_labels · add_label_to_thread · remove_label_from_thread · toggle_labels

Contacts (3 tools)

search_contacts · get_contact_conversations · create_contact

Knowledge Base (5 tools)

list_articles · get_article · create_article · update_article · search_knowledge

Analytics (4 tools)

get_response_times · get_avg_response_time · get_daily_activity · get_closed_activity

Automations (3 tools)

list_automations · toggle_automation · toggle_ai_drafts

WhatsApp (4 tools)

send_whatsapp_message · send_whatsapp_template · list_whatsapp_templates · get_whatsapp_conversation

Comments (2 tools)

add_comment · get_comment

Tags (2 tools)

list_tags · add_tag_to_card

Tasks (1 tool)

create_task

47 tools across 12 categories.

Development

git clone https://github.com/nick-timms/drag-mcp-server.git
cd drag-mcp-server
npm install
cp .env.example .env  # add your API key
npm run build
npm start
Secret-scanning pre-commit hook

This is a public repository. Before committing, install the pre-commit hook so gitleaks scans your staged changes for secrets (keys, JWTs, tokens):

pip install pre-commit   # one-time, if you don't have it
pre-commit install       # installs the git hook in this repo

Run it manually against everything at any time:

pre-commit run --all-files

The hook uses .gitleaks.toml. The same scan runs in CI on every push and pull request (.github/workflows/secret-scan.yml), and npm publish runs a tarball guard (scripts/check-tarball.mjs) that blocks the release if the built package contains any known-sensitive term.

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.