One Folio for Every System Your Property Runs
For three decades, the property tech stack grew by adding layers: a PMS, then a channel manager to reconcile it, then an RMS to price it, a CRM to remember the guest, a tasking app for housekeeping, and a point-of-sale rig for check-in. Each layer needs its own integration with every other layer. Agentic tooling collapses all of it.
Folio is an open CLI and MCP client — the reference implementation of ADAPT-HOP, the Hospitality Operations Protocol. One surface an AI agent (or a human) can drive across every property system. The Hospitality Operations Working Group is forming now to build it — and its first pilot is a proposal to Airbnb.
Collapsing the Middleware Layer
The same protocol that lets an AI agent book a room directly — MCP — can do something quieter and arguably larger: it can let one program drive every operational system inside the property. The fragmented API layer that sits between the PMS, the channel manager, the revenue manager, the CRM, the housekeeping app, and the front desk is not a law of nature. It is an accident of how the stack was assembled, one acquisition and one integration at a time.
Today, connecting six systems means maintaining point-to-point integrations between each pair that needs to talk — brittle webhooks, nightly syncs, and reconciliation jobs that fail quietly at 7pm on a Friday. Every new vendor multiplies the surface area. The channel manager exists almost entirely to paper over this; it is a tax you pay for the privilege of your own systems disagreeing.
ADAPT-HOP replaces that web with a single shared vocabulary. Each system maps once to the protocol instead of N times to its neighbors. An agent — or the folio CLI — speaks that one vocabulary and reaches all of them. The middleware does not get cheaper. It disappears.
Each system maps once to the protocol — not N times to its neighbors. The integration count drops from a web to a spoke.
The Six Layers Folio Converges
None of these systems is wrong. Each solved a real problem. The waste is in the seams between them — and the seams are exactly what an agent-native surface removes.
PMS
Owns rooms, rates, and reservations — but speaks a proprietary dialect every other system must integrate against one by one.
Channel Manager
A translation layer whose only job is to keep Airbnb, Booking, and the PMS from disagreeing. Its existence is a symptom of the fragmentation, not a cure.
RMS
Sets dynamic pricing, then pushes it back through the channel manager and hopes the PMS agrees. Latency and drift are constant.
CRM
Holds the guest profile — duplicated, partial, and out of sync with the booking record the moment a reservation changes.
Ops & Tasks
Turnovers, work orders, and inspections live in a separate app the PMS pings over yet another brittle webhook — if at all.
Front Desk
Identity, payment, and a door code — the moment of arrival — bolted on through point-of-sale hardware and manual workarounds.
Meet folio
An open, vendor-neutral CLI and MCP server. It is named for the guest folio — the single account every department in a hotel already posts to. One folio for the whole property: one command, or one tool an AI agent calls, that reaches every system through ADAPT-HOP.
Why ~8% Is Rational for Airbnb
A reduced fee for connected hosts is not a discount Airbnb absorbs — it is a saving Airbnb earns and shares. Six places the cost comes out:
Lower cost-to-serve
A host-only fee is priced for the cost of supporting a manual, unstructured listing. A folio-connected listing arrives clean, automated, and self-healing — so it costs a fraction to serve. The fee can follow the cost down.
Machine-clean inventory
ADAPT-HOP delivers availability, rates, and restrictions as structured, validated data — no scraping, no rate drift, no reconciliation tickets. Higher-quality supply at lower operational overhead.
Fewer disputes, settled openly
Folio listings resolve disputes through ADAPT-DRP and the Dispute Resolution Working Group — bonded guests, localized arbiters, programmable escrow — taking cases off Airbnb's own resolution center.
Instant programmable settlement
Booking terms, cancellation policy, and payout conditions are encoded into the transaction and clear instantly. No virtual-card overhead, no 30-day reconciliation, no chargeback exposure.
Professional supply, retained
The 15.5% shift is alienating exactly the operators Airbnb most wants — multi-unit, software-run, reliable. A connected tier gives them a reason to stay and to grow inventory on-platform.
Coopetition, not subsidy
ADAPT is not asking Airbnb to discount out of goodwill. It is showing Airbnb how to remove cost from its own books and share part of the saving — preserving margin on a lower cost base.
Commercial Building OS host operators run the reference implementation. At The Exchange Building — 200+ units across 19 floors in downtown Memphis — folio drives the PMS, channel sync, housekeeping, access control, and settlement from one surface. That is where the cost-to-serve numbers behind Folio Direct get measured in production, not modeled on a slide.
From Problem Statement to Pilot
Charter & Landscape
Phase 0- Form HO-WG (ADAPT-WG-005); adopt charter and tri-chairs
- Map the current middleware stack and per-integration cost across PMS, channel managers, RMS, CRM, ops, and front desk
- Problem Statement v0.1: the N×N integration tax and the case for one surface
ADAPT-HOP Draft
Phase 1- Define the shared verb set and schema (availability, rates, reservations, profiles, tasks, access, settlement)
- Publish the folio reference CLI + MCP server as open source
- Working Draft v0.5 — complete operations protocol open for comment
Folio Direct Economics
Phase 2- Model cost-to-serve for a folio-connected listing vs a manual one
- Publish the Folio Direct economic memo — the case for an ~8% connected-host fee
- Open the proposal to Airbnb and the broader platform ecosystem
Pilot at the Exchange
Phase 3- Run folio in production across 200+ units at The Exchange Building via Building OS
- Instrument support load, data quality, dispute rate, and settlement latency
- Pilot report and ADAPT-HOP v1.0 submitted for ratification
Who We Need at the Table
This is a new medium of connectivity, and it will be shaped by the people who show up to define it. The protocol is only as good as the systems and operators in the room.
The teams who own host economics and software-connectivity policy. Folio Direct is a proposal to lower your cost-to-serve — we want it pressure-tested by the people who price it.
The systems folio converges. ADAPT-HOP only matters if the property layer speaks it natively — your implementation feedback shapes the schema.
Today's translation layer. Help define what a single surface absorbs, what it keeps, and how distribution and pricing ride on open rails.
Multi-unit, software-run hosts carrying the 15.5% shift firsthand. You are the ground truth on what the middleware actually costs to run.
Experts in escrow mechanics and programmable settlement — the layer that lets a folio booking clear instantly with terms enforced by protocol.
Teams building travel and property agents that need one control surface to drive instead of six brittle integrations. ADAPT-HOP is built for you to call.
Help Shape This New Medium of Connectivity
The Hospitality Operations Working Group is exploratory and forming now. No fees, no approval needed. Bring a system, an operation, or a hard question — and help decide what one surface should be.