Manage reservations, guests, room types, rates and properties on the Cloudbeds hotel/property-management platform.
Cloudbeds ships in the w6w first-party pack. It declares 13 actions, 2 health checks, and the host runs its code in a sandbox that never sees the credential.
io.w6w.cloudbedsCloudbeds is a hotel and property-management platform: the system a property runs its front desk, reservations, rates and guest records on. This app covers the core of the PMS API v1.3 that a workflow actually needs — searching and creating reservations, checking room-type availability and rate plans for a date range, reading and updating guest records, and reading a property’s own details and daily dashboard.
Two vendor quirks shape the client rather than being left as surprises: some failures answer HTTP 200 with {success: false, message} instead of a 4xx, and the real API base and OAuth authorize/token URLs differ from what a quick guess would produce. Health checks track Cloudbeds’ own status page, scoped to the one component (“Property Management System”) this app’s actions actually depend on, so an unrelated marketing-connector incident elsewhere on the page never drags this app’s health down with it.
The full Cloudbeds surface is far larger than this app covers — allotment blocks, payments and items, housekeeping, webhooks, accounting/fiscal documents, the Insights datasets API, and the channel-manager callback operations are all out of scope for this first pass. See the README for what’s not yet covered.
Three routes to the same 13 actions. The Workflow tab is generated from Cloudbeds'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.
guest-list Search guests by name, contact details, stay window or reservation status. Every filter is optional.
guest-update Update a guest's contact, address, document or company details. At least one field must be supplied.
rate-plan-list List the rate plans for a date range, with their rates and availability.
reservation-create Create a reservation for a property, with rooms, occupancy and guest details.
reservation-list Search reservations by status, date window, room, guest or source. Every filter is optional.
reservation-update Update a reservation's status, arrival time, rooms or custom fields.
room-type-availability-list List the room types available for a date range and occupancy, with their rates.
room-type-list List room types, optionally with their rates for a date range and occupancy.
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 Cloudbeds's real ids.
{
"manifestVersion": "2",
"name": "cloudbeds-example",
"steps": [
{
"id": "guest-update",
"uses": {
"app": "io.w6w.cloudbeds",
"action": "guest-update",
"connection": "conn_YOUR_CONNECTION_ID"
},
"with": {
"guestID": "<guestID>"
}
}
]
}guest-update dashboard-get guest-get guest-list hotel-get +8 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 Cloudbeds, 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 Cloudbeds. 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: "guest-update",
payload: {
guestID: "<value>",
// propertyID: "<value>",
// guestFirstName: "<value>",
// guestLastName: "<value>",
// guestEmail: "<value>",
// guestPhone: "<value>",
// guestCellPhone: "<value>",
// guestGender: "<value>",
// guestAddress1: "<value>",
// guestAddress2: "<value>",
// guestCity: "<value>",
// guestState: "<value>",
// guestZip: "<value>",
// guestCountry: "<value>",
// guestNationality: "<value>",
// guestBirthDate: "<value>",
// guestDocumentType: "<value>",
// guestDocumentNumber: "<value>",
// guestDocumentIssueDate: "<value>",
// guestDocumentIssuingCountry: "<value>",
// guestDocumentExpirationDate: "<value>",
// guestCustomFields: "<value>",
// guestRequirements: "<value>",
// guestCompanyName: "<value>",
// guestCompanyTaxId: "<value>",
// guestTaxId: "<value>",
},
});
if (isActionRun(envelope)) console.log(envelope.value); npm install -g @w6w/cli w6w run conn_YOUR_CONNECTION_ID --action guest-update --payload '{"guestID":"<value>"}' Give an AI agent Cloudbeds — without giving it Cloudbeds'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.cloudbeds#guest-update",
"input": {
"guestID": "<guestID>"
}
}
}
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 Cloudbeds 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.
Cloudbeds'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. Cloudbeds itself is MIT, and the runtime that executes it is source-available (FSL).
Cloudbeds declares its own checks, so its health is a property of the app rather than something the host guesses at.
Component status from status.cloudbeds.com. The verdict comes from the Property Management System component, which is the API this app calls; the other nine components on the page are reported as detail only.