Wardrowbe — ИИ-гардероб на своём сервере

★ 713

Wardrowbe is an open-source self-hosted wardrobe management tool with AI-powered outfit recommendations. Reach for it when your closet is full but you keep wearing the same few things, when you want to audit what you own before buying more, when you need to plan outfits ahead rather than scrambling in the morning, or when you refuse to upload photos of your clothes to a third-party cloud service. Upload photos of each garment, let the AI automatically extract clothing details and apply tags, then receive daily outfit suggestions matched to weather and occasion. The backend is built on Python/FastAPI, the frontend on Next.js/TypeScript, and everything runs via Docker Compose with pre-built multi-arch images for linux/amd64 and linux/arm64 — including Raspberry Pi. Any OpenAI-compatible AI provider works: Ollama for free local inference, OpenAI, LocalAI, or others. Additional features include wear and wash tracking, wardrobe analytics, multi-member household support, scheduled notifications via ntfy/Mattermost/email, and a UI available in 8 languages. All data stays on your own hardware, which sets it apart from mobile wardrobe apps that sync to vendor clouds. Requirements: Docker, Docker Compose, at least 4 GB RAM, and a configured AI provider. Tagging quality depends directly on photo quality — poorly lit or cropped images produce weaker recommendations.