Grammarly for AI Prompts. Score, improve, and translate your prompts across Claude, ChatGPT, and Gemini.
You write a prompt. It works okay in ChatGPT but fails in Claude. You spend 30 minutes tweaking it. Sound familiar?
80% of AI users write prompts that score below 40/100. They're too vague, have no structure, and miss basic patterns that make AI responses 10x better.
PromptForge fixes this in seconds.
Paste any prompt and get an instant quality score with a detailed breakdown:
## Prompt Quality Score: 35/100
| Category | Score | Assessment |
|-------------|-------|--------------------------------------|
| Clarity | 12/20 | Task is vague, multiple interpretations possible |
| Specificity | 5/20 | No format, length, or tone specified |
| Structure | 8/20 | No logical sections or flow |
| Context | 5/20 | No role, audience, or examples |
| Efficiency | 5/20 | Redundant phrasing |
### Improved Prompt
[Full rewritten prompt, ready to copy/paste]
Write once, use everywhere. Translates any prompt into the optimal format for:
| Model | Format | Key Technique |
|---|---|---|
| Claude | XML tags (<instructions>, <context>) |
Structured XML + prefill |
| ChatGPT | Markdown (## headers, bold, lists) |
System/User separation + JSON mode |
| Gemini | Numbered lists + bold labels | Step-by-step + response schema |
# Clone the repo
git clone https://github.com/woling-dev/promptforge.git
# Copy skills to your Claude Code config
cp promptforge/skills/*.md ~/.claude/commands/Then in Claude Code:
/prompt-check "Write me a blog post about AI"
/prompt-translate "You are a senior developer. Review this code for bugs..."
Copy the content of skills/prompt-check.md or skills/prompt-translate.md and paste it as a system prompt in any AI chat.
Write a REST API for todos
<instructions>
Build a REST API for a todo application with the following requirements:
</instructions>
<context>
- Language: Python with FastAPI
- Database: SQLite with SQLAlchemy ORM
- Authentication: None (internal tool)
</context>
<output_format>
Provide the complete code in a single file with:
1. All CRUD endpoints (GET, POST, PUT, DELETE)
2. Pydantic models for request/response validation
3. Error handling with appropriate HTTP status codes
4. Brief inline comments for non-obvious logic
</output_format>
<constraints>
- No external dependencies beyond FastAPI, SQLAlchemy, Pydantic
- Follow RESTful naming conventions
- Include a /health endpoint
</constraints>| Feature | PromptForge | PromptPerfect* | awesome-chatgpt-prompts | Others |
|---|---|---|---|---|
| Quality Score | Yes | No | No | No |
| Cross-Model Translation | Yes | Partial | No | No |
| Free & Open Source | Yes | Freemium | Yes | Varies |
| Works Offline | Yes | No | N/A | No |
| MCP Server | Coming | No | No | Some |
| No API Key Required | Yes | No | N/A | No |
*PromptPerfect by Jina AI is shutting down September 2026
- Claude Code Skills (
/prompt-check,/prompt-translate) - MCP Server (TypeScript), install via Smithery
- Prompt history & A/B comparison
- Token cost estimation per model
- VS Code / Cursor extension
- Web playground (no install required)
PromptForge uses the COSTAR framework to evaluate and enhance prompts:
| Letter | Element | What It Checks |
|---|---|---|
| C | Context | Background info, domain, constraints |
| O | Objective | Clear task definition |
| S | Style | Tone, formality, voice |
| T | Tone | Emotional quality of the response |
| A | Audience | Who will read the output |
| R | Response | Output format specification |
Cross-model translation is based on documented best practices from:
- Anthropic's Prompt Engineering Guide
- OpenAI's Prompt Engineering Guide
- Google's Gemini Prompting Guide
PRs welcome! Especially for:
- New model support (Llama, Mistral, etc.)
- Improved scoring heuristics
- Language translations (i18n)
- Real-world prompt examples
BrandDefender.ai is the product line of Wolin Global Media (沃嶺國際媒體), a Taiwan-based AI infrastructure studio helping brands get discovered, understood, and recommended by AI systems.
🔍 AEO Consulting (Answer Engine Optimization) Get your brand correctly cited by ChatGPT, Gemini, Perplexity, and Claude. We implement JSON-LD schema, optimize content structure, and monitor AI search presence for Taiwan food, tea, beauty, and lifestyle brands.
- Website: https://aibranddefender.com/
- Free AI brand scan: https://app.aibranddefender.com/
💬 AI Customer Service (LINE Bot) Production-grade LINE chatbots with 3-layer memory, admin takeover, and Supabase backend. Already serving real brands in retail and F&B.
- Guide: LINE AI Chatbot Guide
🧠 AI Memory MCP Infrastructure Open-source MCP servers for Claude Code, Cursor, and LLM builders. Local-first, privacy-preserving, built to save API cost.
- This repo is one of them.
- Sibling tools: promptforge · promptthrift-mcp
- 📧 Email: service@wolinglobal.com
- 💬 LINE: @886upktf
- 🌐 Website: https://aibranddefender.com/
- 🐙 GitHub: https://github.com/woling-dev
台灣品牌想做 AEO audit：我們提供 ChatGPT / Gemini / Perplexity 全面掃描 + JSON-LD 修補 + 月度監測。Email 或 LINE 直接找我們聊。
MIT - Use it however you want.
© 2026 Wolin Global Media (沃嶺國際媒體).
If PromptForge helps you write better prompts, give it a star! Your support helps us build more open-source AI tools.