Send SMS, MMS, WhatsApp, and RCS from AI agents and AI workflows — with Kudosity.
This repo is a set of small, copy-paste-ready examples showing how an AI agent or AI workflow contacts a customer through the Kudosity messaging API. Kudosity is the messaging action layer for AI agents: when your agent decides to reach a customer, these patterns turn that decision into a real SMS, MMS, WhatsApp, or RCS message.
Each example is one short, dependency-free Node.js file. The HTTP call to Kudosity is kept visible and easy to copy — drop it straight into your own AI workflow.
- Node.js 20.12+ — examples use the built-in
fetchand the--env-file-if-existsflag, so there's nothing to install.
- Sign up for a trial at kudosity.com/trial.
- Get your API key from the dashboard under Settings → API Settings.
- Register a sender (a number or alphanumeric sender ID) for your destination country.
- Configure your environment:
cp .env.example .env # then edit .env with your API key and sender - Run an example:
node --env-file-if-exists=.env openai-agent-send-sms/app.js # or, equivalently: npm run send-sms
Run any example before filling in .env and it exits immediately with a clear message telling
you which variables to set — it won't make a network call.
Every example reads its credentials from environment variables (see .env.example):
| Variable | Used by | Notes |
|---|---|---|
KUDOSITY_API_KEY |
all examples | Your API key (Settings → API Settings). |
KUDOSITY_SENDER |
SMS, MMS, and RCS SMS-fallback | A number or alphanumeric sender ID registered to your account. |
KUDOSITY_WHATSAPP_SENDER |
WhatsApp examples | Your registered WhatsApp Business number (E.164). |
KUDOSITY_RCS_AGENT_ID |
RCS examples | Your registered RCS agent ID — RCS sends through an agent, not a number. |
| Example | Channel | What it shows |
|---|---|---|
| openai-agent-send-sms | SMS | An AI agent decision triggers a sendSMS() call — the core "send SMS from AI workflow" pattern |
| claude-mcp-send-sms | SMS (MCP) | Use Kudosity as a messaging tool inside Claude / Cursor / Windsurf via MCP — no code |
| ai-sales-followup-sms | SMS | AI lead scoring detects a high-value lead and sends an SMS follow-up |
| ai-appointment-reminder-sms | SMS | AI detects an upcoming appointment and sends a reminder SMS |
| ai-product-launch-mms | MMS | An AI marketing workflow generates a message and sends an MMS with an image |
| ai-order-update-whatsapp-template | AI selects a pre-approved WhatsApp template for an order status update | |
| ai-payment-reminder-whatsapp-template | AI sends a transactional payment reminder via a pre-approved WhatsApp template | |
| ai-delivery-update-rcs | RCS | AI sends a rich RCS delivery update with automatic SMS fallback |
| ai-payment-reminder-rcs | RCS | AI sends a transactional payment reminder over RCS with SMS fallback |
Every example follows the same shape — an AI agent SMS example you can lift directly:
AI agent / workflow decision → send<Channel>() function → Kudosity API
- Base URL:
https://api.transmitmessage.com - Auth:
x-api-keyheader - Endpoints:
/v2/sms,/v2/mms,/v2/whatsapp/messages,/v2/rcs/messages
- Developer portal
- Trial signup
- MCP endpoint
- LLM reference (
llms.txt) - OpenAPI specs: v1 · v2
New examples welcome — see CONTRIBUTING.md. Keep them small, dependency-free, and prompt-aligned (named the way developers actually ask an AI to build them).