Skip to content
MCP server

WACrew

Lets coding agents read, search and reply to WhatsApp chats at a human pace

Runs locallyStatus and history
Pricing
Open source
Platforms
whatsapp, self-hosted, CLI
Runs
Locally
Version
0.1.1
Listing updated
3 days ago

About WACrew

WACrew is an open-source MCP server for communication. It runs locally on your machine and is started by your AI client.

WACrew's MCP package connects an agent to a WACrew panel, the open-source collector that you run yourself and link to WhatsApp by scanning a QR code. From there an agent can list groups, read and search messages, get a daily summary, transcribe voice notes and inspect images, videos and documents.

Sending is deliberately gated. Each chat is either in confirm mode, the default, where the agent shows the text and waits for your OK, or autonomous. Drafts pass a deterministic review that blocks robot-sounding text, and delivery imitates a read receipt, typing indicator and pauses, cancelling itself if the client writes again.

It uses Baileys, the WhatsApp Web multi-device library, not the official API, and the README says to use it only with people who already talk to you, never for bulk messaging. The project is Apache 2.0, with 36 tools and a Portuguese site.

Install

Run in a terminal

claude mcp add wacrew --env WAC_CLOUD_URL='<WAC_CLOUD_URL>' --env WAC_CLOUD_USER='<WAC_CLOUD_USER>' --env WAC_CLOUD_PASS='<WAC_CLOUD_PASS>' -- npx -y wacrew-mcp

Add to claude_desktop_config.json

{
  "mcpServers": {
    "wacrew": {
      "command": "npx",
      "args": [
        "-y",
        "wacrew-mcp"
      ],
      "env": {
        "WAC_CLOUD_URL": "<WAC_CLOUD_URL>",
        "WAC_CLOUD_USER": "<WAC_CLOUD_USER>",
        "WAC_CLOUD_PASS": "<WAC_CLOUD_PASS>"
      }
    }
  }
}

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "wacrew": {
      "command": "npx",
      "args": [
        "-y",
        "wacrew-mcp"
      ],
      "env": {
        "WAC_CLOUD_URL": "<WAC_CLOUD_URL>",
        "WAC_CLOUD_USER": "<WAC_CLOUD_USER>",
        "WAC_CLOUD_PASS": "<WAC_CLOUD_PASS>"
      }
    }
  }
}

Add to .vscode/mcp.json

{
  "servers": {
    "wacrew": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "wacrew-mcp"
      ],
      "env": {
        "WAC_CLOUD_URL": "<WAC_CLOUD_URL>",
        "WAC_CLOUD_USER": "<WAC_CLOUD_USER>",
        "WAC_CLOUD_PASS": "<WAC_CLOUD_PASS>"
      }
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "wacrew": {
      "command": "npx",
      "args": [
        "-y",
        "wacrew-mcp"
      ],
      "env": {
        "WAC_CLOUD_URL": "<WAC_CLOUD_URL>",
        "WAC_CLOUD_USER": "<WAC_CLOUD_USER>",
        "WAC_CLOUD_PASS": "<WAC_CLOUD_PASS>"
      }
    }
  }
}

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "wacrew": {
      "command": "npx",
      "args": [
        "-y",
        "wacrew-mcp"
      ],
      "env": {
        "WAC_CLOUD_URL": "<WAC_CLOUD_URL>",
        "WAC_CLOUD_USER": "<WAC_CLOUD_USER>",
        "WAC_CLOUD_PASS": "<WAC_CLOUD_PASS>"
      }
    }
  }
}

Configuration

  • WAC_CLOUD_URLURL of your WACrew panel (the self-hosted collector).Required
  • WAC_CLOUD_USERPanel login (PANEL_USER).Required
  • WAC_CLOUD_PASSPanel password (PANEL_PASS).Required

Tools this server provides (13)

Taken from the project documentation; the running server may differ.

listar_grupos
List WhatsApp groups
ler_mensagens
Read messages
buscar
Search messages
resumo_do_dia
Summarise the day
transcrever
Transcribe voice notes
ver_imagem
Inspect an image
assistir_video
Inspect a video
ler_documento
Read a document
preparar_resposta
Prepare a reply draft
revisar_texto
Review text for robotic phrasing
responder
Send a reply
status_envio
Check delivery status
cancelar_envio
Cancel a pending send

What it does

Chat reading and search
List groups, read messages and search history.
Voice note transcription
Transcribes audio and inspects images, video and PDFs.
Confirm-first replies
Default mode shows the draft and waits for approval.
Human-paced delivery
Read receipt, typing indicator and pauses.
Robot-text review
Deterministic checks block stock phrases and repeats.
Self-hosted collector
You run the panel with Docker.

Good for

  • Triage a queue of client WhatsApp groups
  • Summarise the day's conversations
  • Transcribe a long voice note
  • Draft replies in your own style for approval

Pricing

Open source

Apache 2.0, no licence fee

Prices change. Check wacrew.com.br before you buy.

WACrew FAQ

Does it send messages automatically?

Only if you set a chat to autonomous; the default asks for your approval first.

Does it use the official WhatsApp API?

No. It uses Baileys; the README warns against bulk messaging.

What do I need?

A running WACrew panel and its URL, username and password in the environment.

What licence is it?

Apache 2.0.

Is WACrew free?

Yes. WACrew is open source, so you can use and self-host it for free.

How do I add WACrew to Claude Code?

Run this command in your terminal: claude mcp add wacrew --env WAC_CLOUD_URL='<WAC_CLOUD_URL>' --env WAC_CLOUD_USER='<WAC_CLOUD_USER>' --env WAC_CLOUD_PASS='<WAC_CLOUD_PASS>' -- npx -y wacrew-mcp

What are the best alternatives to WACrew?

The closest alternatives to WACrew are Ha MCP, Concord MCP, note1, BoardAgent and Callout.

Alternatives to WACrew

  1. Ha MCP

    MCP server

    MCP server for managing and automating Home Assistant from AI assistants

    Runs locally
    • 8 tools
    • Open source
    • 7.9K GitHub stars
    • Productivity
  2. Concord MCP

    MCP server

    Local-first MCP server for messaging and shared work-state between coding agents

    Runs locally
    • Freemium
    • 1.3K GitHub stars
    • Developer tools
  3. note1

    MCP server

    MCP server giving AI assistants search and summaries across your note1 meetings

    • 4 tools
    • Remote
    • Free trial
    • 1 GitHub star
    • Productivity
  4. BoardAgent

    MCP server

    Remote MCP server for a message board where only AI agents post, reply and vote

    • 27 tools
    • Remote
    • Free
    • Community directories
  5. Callout

    MCP server

    Let an AI agent place real phone calls in Japanese or English and read the transcript

    • 5 tools
    • Remote
    • Freemium
    • Communication
  6. CalderFlow

    MCP server

    Turn notes or specs into editable flowcharts, timelines and funnels from an AI assistant

    • 10 tools
    • Remote
    • Free
    • Design and media