Send WhatsApp, SMS, email and social messages, and manage contacts, conversations, notes, templates and webhooks in Superchat.
Superchat ships in the w6w first-party pack. It declares 53 actions, 3 health checks, and the host runs its code in a sandbox that never sees the credential.
io.w6w.superchatSuperchat is a shared inbox for customer messaging — WhatsApp, SMS, email, Instagram, Facebook Messenger, Telegram and a web chat widget in one workspace — and this app drives it from a workflow through one API key.
Its centre is a single Send Message action that works across every connected channel: pass a channel, a recipient (a contact, an E.164 phone number or an email address; an unknown one creates the contact) and a content type — text, media, email, generic or WhatsApp template, quick reply or list. Templates, template folders, files and message delivery analytics are covered, so a workflow can start a WhatsApp conversation with an approved template and later read whether it was delivered, read or clicked.
The rest is the CRM around those messages: create, search, update and delete contacts and custom attributes, manage contact lists, read and update conversations (status, assignee, labels, snooze), keep internal notes, export a conversation, and manage the webhook subscriptions that report inbound messages and contact changes back out. Health checks cover Superchat’s own status page and API reachability; webhook signing secrets are redacted from every read so they do not land in run records.
Three routes to the same 53 actions. The Workflow tab is generated from Superchat's own manifest and carries its real ids, so it is copy-pasteable; the Code and CLI examples are the same call for any action on any app, so every app-specific value in them is a blank you fill in.
channel-list List channels (the senders you can pass to Send Message), newest first, one cursor page at a time.
contact-contact-lists-list List the contact lists a contact belongs to, newest first, one cursor page at a time.
contact-conversations-list List a contact's conversations, newest first, one cursor page at a time.
contact-create Create a contact. At least one email address or phone number (E.164) is required, because the API requires a handle.
contact-remove-from-list Remove a contact from a contact list (the contact itself is kept).
contact-search Find contacts by an exact email, phone, Instagram username or custom-attribute value. The API accepts exactly one search expression per call.
contact-update Update a contact's name, gender, handles or custom attributes. The API's PATCH requires first name, last name and gender on every call (a missing one would be read as a clear), so this action reads the contact first and re-sends the values you did not change.
conversation-export-create Start an export of a conversation over a time range. Poll Get Conversation Export for the download link.
conversation-export-get Check an export's status; when `done`, `link.url` is a time-limited download URL.
conversation-messages-list List the messages in a conversation, optionally only one direction or a time window.
conversation-update Change a conversation's status, inbox, snooze time, assigned users or labels. Only the fields you set are sent.
custom-attribute-create Define a new custom contact attribute (text, number, date, date-time, single-select or multi-select).
custom-attribute-list List the custom contact attributes defined in the workspace.
message-analytics-get Delivery and engagement timestamps (attempted, delivered, failed, read, clicked, replied) for up to 100 messages.
message-send Send a message on one of your channels (WhatsApp, SMS, email, Instagram, Messenger, Telegram) to a contact id, phone number (E.164) or email address. An unknown phone/email creates the contact.
note-create Add an internal note (never sent to the contact) to a conversation.
template-analytics-get Aggregated send, delivery, read, click and reply counts and rates per template over a period.
template-folder-create Create a template folder, optionally nested or tied to a WhatsApp Business Account.
template-folder-list List template folders, newest first, one cursor page at a time.
webhook-create Subscribe an HTTPS URL to Superchat events. The response carries the signing `secret` ONCE here; store it now, because every other read of the subscription is redacted by this app.
webhook-get Fetch one webhook subscription. The signing `secret` is redacted.
webhook-list List webhook subscriptions. The signing `secret` is redacted.
webhook-update Replace a subscription's target URL and/or events (PUT). Events you leave out are not kept: pass the full list you want. The signing `secret` is redacted.
A workflow step names the app and the action, and the editor fills in the
connection when you pick one. This is the Step shape from the
workflow spec, carrying Superchat's real ids.
{
"manifestVersion": "2",
"name": "superchat-example",
"steps": [
{
"id": "channel-get",
"uses": {
"app": "io.w6w.superchat",
"action": "channel-get",
"connection": "conn_YOUR_CONNECTION_ID"
},
"with": {
"channelId": "<channelId>"
}
}
]
}channel-get channel-list contact-add-to-list contact-contact-lists-list contact-conversations-list +48 more actions available
Every app-specific value here is a blank you have to fill in. An
app action is reached through the connection that authenticates it, so the
address is a connection id, not the app id — and connections belong to your account,
so a public page cannot know yours. Create one for Superchat, then fill in
the three blanks: conn_YOUR_CONNECTION_ID, the action key, and the
parameters that action declares. The call itself is real — the shape is transcribed
from the studio's own snippet builder, which prints the same kind of blanks — but
nothing in it is specific to Superchat. The Workflow tab is where this app's
real ids are.
npm install @w6w/sdkyarn add @w6w/sdkpnpm add @w6w/sdkdeno add npm:@w6w/sdkimport { W6wClient, isActionRun } from "@w6w/sdk";
// Reads W6W_BASE_URL and W6W_TOKEN from the environment when omitted.
const client = new W6wClient();
const envelope = await client.run({
urn: "conn_YOUR_CONNECTION_ID",
action: "channel-get",
payload: {
channelId: "<value>",
},
});
if (isActionRun(envelope)) console.log(envelope.value); npm install -g @w6w/cli w6w run conn_YOUR_CONNECTION_ID --action channel-get --payload '{"channelId":"<value>"}' Give an AI agent Superchat — without giving it Superchat's credentials. One MCP endpoint exposes every app, function and workflow the caller is entitled to, as tools it can discover and run. Access is granted per team while we onboard.
One tool call{
"name": "w6w_invoke",
"arguments": {
"ref": "app:io.w6w.superchat#channel-get",
"input": {
"channelId": "<channelId>"
}
}
}
Every tool names its target with a single ref. The
app: form above doesn't name a connection at all — the
host resolves which of the caller's Superchat connections to sign
with, and refuses rather than guesses when the answer is ambiguous.
The token is attached host-side, at the moment of the call. It is never a tool argument, never in the model's context, and never in a transcript — so a prompt injection has nothing to exfiltrate.
Tools are derived per end user from what that person has actually connected and is entitled to — not one shared bot identity carrying the union of everyone's access.
Multi-step work runs on the workflow engine and returns a run handle the agent can poll — retries, branching and state survive the conversation that started them.
Superchat's declared health checks are on the surface too, so an agent can tell "the vendor is down" from "your credential expired" before it burns a retry on either.
The MCP surface is part of the hosted platform. Superchat itself is MIT, and the runtime that executes it is source-available (FSL).
Superchat declares its own checks, so its health is a property of the app rather than something the host guesses at.
Resource status from status.superchat.de (Better Stack): WebApp, API, Webchat, Worker, Webhook Infrastructure and the review page.
Unauthenticated GET /v1.0/me. The auth layer's empty 401 passes: it proves the API and its gateway are answering. Key validity is the derived `auth:*` check's job.
Not exposed: the limit is 2500 requests per 5 minutes per workspace, but no header or endpoint reports usage.