Supports validating recipients on Airtel's network and sending live test RCS messages via its RBM API.
Supports validating recipients on AT&T's network and sending live test RCS messages via its RBM API.
Supports validating recipients on Deutsche Telekom's network and sending live test RCS messages via its RBM API.
Supports validating recipients on Jio's network and sending live test RCS messages via its RBM API.
Supports validating recipients on Movistar's network and sending live test RCS messages via its RBM API.
Supports validating recipients on Orange's network and sending live test RCS messages via its RBM API.
Supports validating recipients on Verizon's network and sending live test RCS messages via its RBM API.
Supports validating recipients on Vodafone's network and sending live test RCS messages via its RBM API.
This server does not currently have a deployable release on Glama. Follow the repository's setup instructions to install and configure it for your MCP client, including any required credentials and local dependencies.
An MCP (Model Context Protocol) server that lets any AI assistant — ChatGPT, Claude, Cursor, Cline, Continue, Cody — compose, preview, validate, and send RCS Business Messages from inside the chat.
Tool | What it does |
| Draft an RCS rich card. Required: |
| Render the message as an interactive iPhone widget inside the chat. The widget lets the user edit and push to a real iPhone. |
| Check the recipient against an 80+ MNO database (AT&T, T-Mobile, Verizon, Dish, US Cellular, O2 UK, Vodafone, EE, Deutsche Telekom, Jio, Airtel, Singtel, Orange, SFR, Bouygues, Movistar, …). Returns PASS/WARN with carrier-specific warnings + recommendations. |
| Push the message to the RCS X iOS app via APNs for final visual proof before live send. Destructive (pushes to a real device). |
| Send a live test RCS via the MNO RBM API. Returns a delivery receipt with MNO name + per-message cost. Destructive + open-world (touches carrier billing). |
Related MCP server: iMessage Max
Path | Purpose |
| MCP JSON-RPC 2.0 endpoint. Stateless — every request is independent. |
| Apps SDK discovery manifest (name, icon, categories, tools). |
| OpenAI Apps submission challenge token. |
| The iPhone preview widget (CSP-protected, served from same origin). |
npm install
npx wrangler dev # local dev server at http://localhost:8787
npm test # 15 integration tests, all greennpx wrangler deploy # publishes to rcsx-mcp-server.<acct>.workers.dev/mcpOpen ChatGPT → Settings → Connectors → Developer mode (toggle on; requires org-admin).
Add connector → URL: https://rcsx-mcp-server.<acct>.workers.dev/mcp → Label: RCS X.
In any chat: "Draft an RCS to +14155551234 promoting a 40% off Black Friday deal."
Edit ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"rcsx": {
"command": "npx",
"args": ["-y", "mcp-remote@latest", "https://rcsx-mcp-server.<acct>.workers.dev/mcp"]
}
}
}Settings → Features → Model Context Protocol → Add new global MCP server:
{
"mcpServers": {
"rcsx": {
"url": "https://rcsx-mcp-server.<acct>.workers.dev/mcp"
}
}
}Cloudflare Workers — single-isolate, edge-deployed, sub-100ms cold start.
Stateless MCP transport — every request is independent; fresh McpServer + fresh WebStandardStreamableHTTPServerTransport per incoming POST (per the SDK's _hasHandledRequest contract).
@modelcontextprotocol/sdk 1.29+ with full inputSchema + outputSchema on every tool (Apps SDK requirement) and proper destructiveHint / openWorldHint annotations.
Zod 4 for schema validation. Tool implementations are pure functions that return MCP-shaped { content, structuredContent, _meta }.
In stateless mode, the MCP SDK requires a fresh Server + Transport per request — both because the transport can only handle one request (_hasHandledRequest flag) AND because a server can only be connected to one transport at a time. Tool registration is cheap (Zod schemas + JSON), so the trade-off favors correctness.
Apache-2.0 — see LICENSE.
This server cannot be deployed
Business texting for AI agents: send texts, read threads, manage contacts and run campaigns.
Run WhatsApp Business campaigns from any AI assistant: contacts, segments, and broadcasts.
Send and receive iMessage, SMS, RCS, and WhatsApp, and manage contacts, chats, and numbers.
Create and edit flows, read responses, and reply to users — from your AI assistant.
