Evaluable
Prompts should not only look good but also verify whether the model actually completes the task.
About
Making AI Outputs More Stable and Verifiable
A structured prompt navigator that helps teams discover, evaluate, and reuse high-quality prompts
Project Overview
2124
Featured Prompts
8
Capability Categories
8
Topic Collections
959
Advanced Scenarios
Core Design Principles
Structured
Each entry has category, difficulty, tags, reference output, or scoring dimensions for easy team reuse.
Iterable
Now upgraded to a database-driven system with admin editing, automated ingestion, and source synchronization. Can further evolve into submissions, reviews, evaluation results, and version history.
Bounded
Scenarios like security, privacy, healthcare, and finance prioritize boundary prompts and alternatives.
Technical Architecture
Next.js 16 + React + TypeScript
Modern full‑stack framework with excellent performance and development experience
Prisma + SQLite
Lightweight database supporting fast queries and full‑text search
FTS5 + Semantic Search
Hybrid keyword and semantic search for quickly finding relevant content
AI SDK + Ollama
Supports multiple models with intelligent prompt processing
Current Scope
Easy Prompt launches as a navigator powered by a database-driven content management system, supporting admin editing, automated ingestion, and source synchronization to continuously accumulate high-quality structured prompts. The long-term direction is to build a model capability testing database, enabling multi-model comparison, version tracking, and scoring reviews.
- No login is required to browse and copy prompts; submitting prompts requires an account.
- No model API integration yet; no online runtime costs.
- Content is maintained via a database with admin editing, automated ingestion, and source synchronization, prioritizing quality and structural stability.
- User submissions and admin moderation are supported; real copy statistics and model comparison data can come next.
Future Plans
2025 Q2: User login, submission, and review system
2025 Q3: Model comparison and multi‑version tracking
2025 Q4: Team collaboration and sharing features
Contribution Guidelines
Welcome to submit your high‑quality prompts and help more developers
Feedback & Suggestions
Have suggestions or found issues? Contact us via GitHub Issues
GitHub: https://github.com/lindakun/easy-prompt