mcpField
W
WasapFlow Bridge

WasapFlow Bridge

Remote
wasapflow.comCommunication0Updated 5d ago

About WasapFlow Bridge

WasapFlow Bridge is a WhatsApp Business API bridge for teams building their own CRM or SaaS, accompanied by a compact Node.js CRM starter. It provides number connection, an inbox, and webhook handling as a base for further development.

Key features

  1. 01Connect numbers through Meta Embedded Signup
  2. 02Receive signed webhooks and message statuses
  3. 03Manage conversations, unread counts, and replies
  4. 04Support multiple numbers and BSUID contacts
  5. 05Guidance for AI coding tools

Use cases

  1. 01Add WhatsApp Business to a custom CRM
  2. 02Review message status and handle conversations
  3. 03Extend the starter with a multi-tenant inbox

MCP support

An optional WasapFlow MCP server lets AI search the documentation and call the Bridge account, for example to list numbers, inspect failed webhooks, or send a test message.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://officialapi.wasapflow.com/bridge/mcp
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

All endpoints
Streamable HTTPhttps://officialapi.wasapflow.com/bridge/mcpAuth required

Tools

Details

From the project README

WhatsApp CRM Starter: official WhatsApp Business API in your own app

A minimal, working WhatsApp CRM on the official WhatsApp Business API (Meta Cloud API) through WasapFlow Bridge. You don't need to become a Meta Tech Provider or go through Meta App Review. Clone it and run it, then let your AI (Claude Code, Cursor, Codex, Copilot) grow it into your CRM or SaaS.

It's about 400 lines of Node.js across four files, with two dependencies and no build step.

What it does

  • Connect WhatsApp button. Your client connects their number through Meta Embedded Signup. Coexistence is supported, so the business can keep using the WhatsApp Business app on their phone.
  • Signed webhooks. Incoming messages, delivery and read status, replies typed on the phone (message.echo), and chat history backfill (message.history).
  • Inbox. Conversations, unread counts, the 24-hour window, and text or template replies.
  • Multi-number. Everything is keyed by phone_number_id.
  • BSUID-safe contacts. WhatsApp usernames don't break your phone field.
  • AGENTS.md, CLAUDE.md and .cursor/rules. These tell your AI exactly how the integration works, so what it builds on top stays correct.

Run it in 5 minutes

  1. Partner account. Create a free account at https://partner.wasapflow.com/bridge/partner/register and save a card. Nothing is charged until the first number connects. After that, Bridge costs $10/month for every 3 WhatsApp numbers, and Meta bills message charges directly. WasapFlow Bridge is only the bridge to Meta: this app is your inbox, and numbers are connected from its own Connect WhatsApp button (the WasapFlow portal has no connect button).
  2. Install:
    git clone https://github.com/kobaranteguh/whatsapp-crm-starter.git
    cd whatsapp-crm-starter
    npm install
    cp .env.example .env
    
  3. Keys. Open portal → Integration and copy the Partner Key and Webhook Secret into .env.
  4. Public URL for webhooks. Deploy the app (Railway, Render, Fly, a VPS), or tunnel your local machine (cloudflared tunnel --url http://localhost:3000 or ngrok http 3000). In portal → Integration, save the webhook URL as https://<your-domain>/webhook/wasapflow.
  5. Start:
    npm start
    
    Open http://localhost:3000 (user admin, password = ADMIN_PASSWORD), click + Connect WhatsApp, then message that number from your phone.

Build the rest with your AI

Open the folder in your AI tool and paste one of these prompts:

Read AGENTS.md and the WasapFlow Bridge API reference. Replace lib/store.js with Postgres using Prisma, keeping the same exports. Add migrations.

Read AGENTS.md. Make this multi-tenant: users sign up, each tenant connects its own WhatsApp numbers via POST /connect/session with state = tenant id, and only sees its own inbox.

Read AGENTS.md and the Bridge docs. Add media: show incoming images and documents, and let agents send images from the composer.

Read AGENTS.md. Add a templates page: list templates with status, create new ones, and refresh on the template.status_updated webhook.

Read AGENTS.md. Add assignment and tags: assign a conversation to a staff member, tag it, and filter the inbox by tag.

Give your AI live access (optional). The WasapFlow MCP server lets it search the docs and call your Bridge account, for example to list numbers, check failed webhooks, or send a test message.

claude mcp add --transport http wasapflow https://officialapi.wasapflow.com/bridge/mcp --header "x-partner-key: wf_..."
{ "mcpServers": { "wasapflow": { "url": "https://officialapi.wasapflow.com/bridge/mcp", "headers": { "x-partner-key": "wf_..." } } } }

Files

FileWhat it is
server.jsWebhook receiver (signature check, events), /api/*, Basic auth
lib/bridge.jsThe only module that calls Bridge (plain fetch, no SDK)
lib/store.jsJSON-file storage that you can replace with your database
public/index.htmlInbox UI
AGENTS.mdRules and recipes for AI coding tools

Links

FAQ

Do I need a Meta Business verification or app review? You don't need Meta App Review or Tech Provider status; Bridge provides both. Each business connects its own number through Meta's Embedded Signup from your app's Connect WhatsApp button and must follow WhatsApp's policies.

Can a business keep using WhatsApp on their phone? Yes. With Coexistence, replies typed on the phone show up in this inbox as message.echo.

Is this production-ready? The integration patterns are: signature check, deduplication, BSUID handling, multi-number, and the 24-hour window. The JSON-file storage and Basic auth are not. Swap them in for your stack (see the prompts above).

MIT licensed. © Kobaran Teguh Sdn. Bhd.

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.