Cooldown

Overview

This repo contains all the components for the cooldown app. This app aims to provide users a way to define actions that have a "cooldown" period, that notifies them when they can re-perform the action. An example would be "watering the flowers - 48 hour cooldown", it is important to note that just because the 48 hour cooldown has completed doesn't mean the user will re-attempt the action in the near future.

Architecture decisions

• Keep features simple first: implement read-only listing before mutations.
• Keep domain and presentation separate in the web app: hooks own fetching/state, components render.
• Use the Blocked UI paradigm for frontend feature composition (blocks, boundaries, and feature groups).
• Use SQLite in the API for fast local persistence and simple deployment.
• Use SQLx for database access and Axum for HTTP routing in Rust.
• Use TanStack Query + TanStack DB in the web app: Query fetches remote data, DB stores/query local collection state.
• Use Radix UI standard components for UI building blocks.
• Prefer explicit typing and helper functions over as casts; treat assertions as a last resort.

Blocked UI docs

Blocked UI reference docs for this repo are in .codex/blocked-ui-docs/.

• Start at .codex/blocked-ui-docs/00-index.md.
• Agent-facing implementation rules are summarized in AGENTS.md.

Structure

.
├── deploy
├── infra
├── api
├── README.md
└── web-app

Production deployment

Production deployment is split into:

• infra/opentofu: Hetzner infrastructure as code (server, firewall, volume, SSH key)
• deploy: Kamal app configs for API and web services

See:

• infra/opentofu/README.md
• deploy/README.md