Loading...
Professional-grade AI prompt engineering tool for microstock image creators
Microstock image creators struggle with scattered prompt workflows across multiple tools and platforms. They need to organize, refine, and reuse prompts efficiently while maintaining offline access and preventing data loss during browser sessions.
Built a professional-grade prompt engineering tool with dual-mode input system (quick entry and detailed editor), offline-first architecture using IndexedDB via Dexie.js, and state management via Zustand. Solved React 19 hydration race conditions between server and client state, implemented robust queue system for prompt operations.
Solo developer — architecture, implementation, and testing. Designed the offline-first data layer, implemented hydration fixes, and built the dual-mode UI system.
Offline-first architecture with zero data loss across browser sessions. Hydration race conditions eliminated through careful state synchronization. Tool handles complex prompt workflows without external dependencies.