English | 中文

Prompt Optimizer 🚀

A professional Claude Code skill that turns simple instructions into production-ready prompts using adversarial evaluation and project context.

Watch how a vague request becomes a structured, edge-case-proof prompt in seconds.

Features

• Context-Aware Optimization: Analyzes your entire project structure (files, design docs) to generate relevant prompts, not generic ones.
• Adversarial Testing: Automatically simulates "Red Teaming" to find logical loopholes in your prompt before you run it.
• Quantitative Evaluation: Detailed scoring system (0-100) based on clarity, specificity, and robustness.
• Interactive Review (WebView): A native desktop UI (Wails) to diff, edit, and confirm changes without leaving your workflow.
• Ready-to-use output: Returns a formatted prompt you can copy directly into Claude.

Installation

One-line install

macOS/Linux:

curl -fsSL https://raw.githubusercontent.com/geq1fan/prompt-optimizer-skill/main/install.sh | bash

Windows (PowerShell):

irm https://raw.githubusercontent.com/geq1fan/prompt-optimizer-skill/main/install.ps1 | iex

Manual install

# Clone into the Claude Code skills directory
git clone https://github.com/geq1fan/prompt-optimizer-skill ~/.claude/skills/prompt-optimizer-skill

Update

# macOS/Linux
~/.claude/skills/prompt-optimizer-skill/install.sh update

# Windows
& "$env:USERPROFILE\.claude\skills\prompt-optimizer-skill\install.ps1" -Action update

Usage

Optimize a prompt

/optimize-prompt Write a function to parse JSON

Iterate

/optimize-prompt iterate Add error handling requirements

How it works

1. Analyze: Checks prompt clarity, completeness, and structure
2. Choose a strategy: Selects an optimization approach based on complexity
3. Enhance: Improves the prompt while preserving intent
4. Evaluate: Provides actionable feedback and a score
5. Interactive Review: Use the WebView app to review and confirm the optimized prompt

Contributing

We welcome contributions! Please see CONTRIBUTING.md for details on how to submit pull requests, report issues, and suggest improvements.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgements

Inspired by linshenkx/prompt-optimizer.