A self-hosted inventory and collection manager with user-defined schemas. Build your own structure for anything — wardrobe, kitchen, board games, tools, whatever — then search, filter, and browse it all from one place.

Open Trove lets you create collections (called Groups), define schemas that describe the shape of items in that collection, then add items with structured data, tags, and images. Think of it like a personal database with a UI that adapts to whatever you're tracking.

A clothing collection might have fields for brand, size, season, and price. A recipe collection might have text areas for ingredients and instructions. A board game shelf might track player count and play time. You define the structure — the app builds the forms and filters to match.

• Groups — Top-level collections (e.g. "Wardrobe", "Kitchen", "Board Games")
• Schemas — Templates that define what fields an item has, organized into sections
• Items — Individual entries within a group, created from a schema
• Tags — Freeform labels on any item for cross-cutting organization
• Images — Multiple images per item with thumbnail generation and reordering

• Schema editor — Visual editor for building item structures. Add sections, name fields, pick types, configure options. Rename anything inline. Change field types after creation with smart data conversion warnings.
• 11 field types — string, textarea, int, float, boolean, datetime, dropdown, multiselect, unit, computed, image
• Unit fields — Built-in unit categories (mass, volume, length, currency, count) with conversion support
• Computed fields — Formula-based fields that reference other fields
• Full-text search — SQLite FTS5 indexes item names, field data, and tags. Searches across everything or scoped to a collection.
• Faceted filtering — Sidebar filters that adapt to your schema: dropdowns for single-select fields, checkboxes for multi-select, min/max range inputs for numeric fields, all with item counts
• Image management — Upload multiple images per item, auto-generated thumbnails, drag to set primary/thumbnail image
• Dark mode — System-aware with manual toggle, persisted to localStorage
• Import/Export — JSON and CSV export per collection or schema, with import support
• Directory Views — Custom hierarchical views of items within a group
• Seed data — Ships with example collections (Wardrobe, Kitchen, Board Games) to try things out

Schemas are JSON documents with a sections object. Each section contains named fields, and each field has a type plus optional configuration.

mass, volume, length, currency, count

{
  "sections": {
    "Basics": {
      "category": {
        "type": "dropdown",
        "options": ["Top", "Bottom", "Outerwear", "Shoes", "Accessory"]
      },
      "brand": {
        "type": "string"
      },
      "size": {
        "type": "dropdown",
        "options": ["XS", "S", "M", "L", "XL"]
      }
    },
    "Details": {
      "season": {
        "type": "multiselect",
        "multiselect-items": ["Spring", "Summer", "Fall", "Winter"]
      },
      "condition": {
        "type": "dropdown",
        "options": ["New", "Like New", "Good", "Fair", "Worn"]
      }
    },
    "Pricing": {
      "purchase_price": {
        "type": "unit",
        "unit_category": "currency",
        "default_unit": "USD"
      },
      "date_acquired": {
        "type": "datetime"
      }
    }
  }
}

You can change a field's type after creation through the schema editor. Compatible conversions (like string → textarea, or int → float) preserve existing data silently. Incompatible changes show a warning since existing item data may not match the new type.

Safe conversions include:

• string ↔ textarea ↔ dropdown
• int ↔ float → string
• boolean → string
• dropdown ↔ multiselect

• Python 3.12+
• Node.js 18+

cd organizer
python -m venv .venv
source .venv/bin/activate
pip install -r backend/requirements.txt
uvicorn backend.main:app --reload

The API runs at http://localhost:8000. SQLite database is created automatically in data/organizer.db.

cd frontend
npm install
npm run dev # add ` -- --host 0.0.0.0` to run on all interfaces

Dev server runs at http://localhost:5173 and proxies API calls to the backend.

With the backend running:

pip install httpx  # if not already installed
python seed_data.py

This creates three example collections with sample items — a Wardrobe with clothing and outfits, a Kitchen with recipes, and a Board Games collection.

backend/
  main.py              # App init, lifespan, router registration
  models.py            # SQLAlchemy models (Group, ItemSchema, Item, etc.)
  database.py          # Async engine + session factory
  routers/
    groups.py          # CRUD for collections
    schemas.py         # CRUD for item schemas
    items.py           # CRUD for items
    images.py          # Upload, serve, thumbnail, reorder
    search.py          # FTS5 search + faceted filtering
    units.py           # Unit conversion API
    views.py           # Directory view management
    export.py          # JSON/CSV import and export
  services/
    search.py          # FTS5 table management, triggers, filtering
frontend/
  src/
    pages/
      GroupsPage.tsx        # Collection list (home)
      GroupDetailPage.tsx   # Items + schemas in a collection
      ItemDetailPage.tsx    # Single item view + edit
      SchemaEditorPage.tsx  # Visual schema builder
      SearchPage.tsx        # Search + faceted filter sidebar
    api.ts                  # API client
    types.ts                # TypeScript interfaces

Personal project. Do whatever you want with it.