CapyInn

Offline-first property management software for mini hotels

A desktop PMS for small hotels and guesthouses in Vietnam.

Property onboarding · Vietnamese ID OCR · Check-in/check-out · Reservations · Night audit

Why CapyInn · Demo · Features · Local development · Verification

Built for mini hotels that need one local app for room status, guest intake, nightly billing, housekeeping, and end-of-day reconciliation.

CapyInn is a desktop app for mini hotels and guesthouses that need a local-first operating tool without relying on a remote backend. The project focuses on real front-desk workflows: room layout setup, faster guest intake, Vietnamese ID OCR, nightly pricing, housekeeping, revenue reporting, and end-of-day reconciliation.

Note: CapyInn is a clean-slate rename from MHM. Current builds use the new runtime root at ~/CapyInn and do not auto-migrate legacy local data from ~/MHM.

Table of contents

• What CapyInn solves
• Product demo
• Key features
• Tech stack
• System requirements
• Local development
• Verification
• Repository layout
• Known limitations
• Additional docs
• Contributing
• License

What CapyInn solves

CapyInn is built for a narrow but practical use case: small hotels that need a system they can run locally, control directly, and adopt without a long setup project.

Before	With CapyInn
Handwritten logs and fragmented tracking	Room status, bookings, and transactions live in one app
Manual guest registration entry	OCR extracts Vietnamese ID details and speeds up intake
Nightly pricing calculated by hand	Check-in, extend-stay, check-out, and folio flows are automated
End-of-day reporting done manually	Dashboard, analytics, expenses, and night audit are built in
Initial setup takes too much time	Onboarding generates room types, layouts, and operating defaults

Product demo

Dashboard

Guest check-in and booking flow

Guest profile and operations

Analytics and settings

Key features

Onboarding and property setup

• Configure hotel identity, check-in and check-out rules, invoice details, and app lock
• Create room types and default pricing during the first-run wizard
• Generate a room layout by floors, room count, and naming scheme

Front-desk operations

• Dashboard organized around the configured room layout
• Check-in, check-out, extend-stay, and reservation flows in one desktop app
• Support for multiple guests on the same booking
• Fast copy flow for guest registration details

Vietnamese ID OCR

• Local OCR powered by PaddleOCR v5 through ocr-rs
• Watches ~/CapyInn/Scans/ for new scan files
• Extracts guest name, national ID number, birth date, and address for check-in

Billing, payments, and reporting

• Night-based pricing by room type
• Charge, payment, deposit, and balance tracking
• Revenue analytics, expense tracking, and CSV export

Housekeeping and night audit

• Post-checkout housekeeping state tracking
• Maintenance notes per room
• Night-audit flow for daily reconciliation

MCP and automation integrations

• CapyInn can be extended through MCP-friendly workflows for operator tooling and agent-driven automations
• These flows can be paired with OpenClaw and n8n for custom orchestration around hotel operations
• For Zalo personal chat automations, you can use the prebuilt community node n8n-nodes-zca-zalo, published on npm and built on top of zca-js

Crash reporting

• Severe crashes are always written locally under ~/CapyInn/diagnostics so the app can recover on the next launch.
• Sending a sanitized crash report to Sentry is optional and controlled in Settings. The remote report flow does not include usage analytics, session replay, guest records, or raw OCR payloads.

Tech stack

Layer	Technology
App shell	Tauri 2
Backend	Rust + SQLite (sqlx)
Frontend	React 19 + TypeScript
State	Zustand
UI	Tailwind CSS 4 + shadcn/ui
OCR	ocr-rs + PaddleOCR v5 + MNN
Charts	Recharts
Tests	Vitest + Rust tests + Clippy

System requirements

Component	Requirement
macOS	12+
Node.js	20+
Rust	stable via rustup
Xcode CLT	recent version
Disk footprint	roughly 25MB before operational data

The project is currently verified most heavily on macOS and Apple Silicon.

Local development

Install prerequisites

curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh
source ~/.cargo/env
xcode-select --install
node --version

Clone and run the desktop app

git clone https://github.com/chuanman2707/CapyInn.git
cd CapyInn/mhm
npm ci
npm run tauri dev

Build a release bundle

cd CapyInn/mhm
npm run tauri build

Release bundles are generated under mhm/src-tauri/target/release/bundle/.

Verification

cd CapyInn/mhm
npm test
npm run build
cargo check --manifest-path src-tauri/Cargo.toml
cargo test --manifest-path src-tauri/Cargo.toml
cargo clippy --manifest-path src-tauri/Cargo.toml --all-targets -- -D warnings

If you only need the web UI during frontend work:

cd CapyInn/mhm
npm run dev

Repository layout

CapyInn/
├── Public/                 # README demo screenshots
├── mhm/
│   ├── src/                # React UI, stores, pages, components
│   ├── src-tauri/          # Rust backend, IPC commands, DB, gateway, OCR
│   ├── tests/              # Vitest suites and mocked desktop flows
│   ├── public/             # Static assets
│   └── models/             # OCR models
├── PRD.md                  # Product requirements
├── CONTRIBUTING.md
├── SECURITY.md
└── README.md

Known limitations

• OCR is currently optimized for Vietnamese national ID cards; passports and international documents are not complete yet
• Windows and Linux are not first-class targets yet
• The project is designed for mini-hotel scale, not large chain operations

Additional docs

• PRD
• Contributing guide
• Security policy
• Changelog

Contributing

Read CONTRIBUTING.md before opening a pull request.

Short checklist:

1. Fork the repository
2. Create a branch from main
3. Keep commit messages in Conventional Commits format
4. Re-run npm test, npm run build, cargo check, cargo test, and cargo clippy
5. Open a pull request with scope and verification notes

License

CapyInn is released under the MIT License.