Prompt Optimizer

Rewrites weak prompts into tight, effective ones using vibe-lint rules + Claude

Features

• Interactive prompt analysis — Scans for 20+ prompt antipatterns using vibe-lint rules
• Real-time feedback — Shows warnings and errors with specific suggestions
• Claude API integration — Automatically rewrites prompts with optimization suggestions
• Color-coded output — Visual severity indicators (errors in red, warnings in yellow)
• CLI-first design — Works directly in terminal with readline interface

Prerequisites

• Node.js 14 or higher
• Claude API access (Anthropic account required)

Installation

npm install -g prompt-optimizer

Or run directly with npx:

npx prompt-optimizer

Configuration

Set your Claude API key as an environment variable:

export ANTHROPIC_API_KEY="your_api_key_here"

Usage

Launch the interactive prompt optimizer:

prompt-optimizer

The tool will analyze your prompts for common issues including:

Error Categories (Will block optimization)

• Ambiguous scope — "do your best", "handle it", "as needed"
• Role confusion — "you are an AI assistant"
• Missing context — References to "the code above" without providing it
• Contradictions — "be brief and comprehensive"

Warning Categories (Suggestions for improvement)

• Vague quality — "make it better", "high-quality"
• Output format — "in a good format", "respond appropriately"
• Politeness bloat — "please try to", "feel free to"
• Negation-only — "don't be vague" without positive guidance

Example

Input prompt:

Please try to make my code better and more professional. Handle any issues as needed and respond in a good format.

The tool identifies:

• W008: "Please try to" (politeness bloat)
• W001: "make it better" (vague quality)
• W003: "more professional" (subjective without rubric)
• E002: "as needed" (ambiguous scope)
• W004: "good format" (unspecified structure)

Optimized version:

Refactor this code to pass ESLint strict rules and achieve <10 cyclomatic complexity. Add JSDoc comments to all public functions. Return as a markdown code block with before/after sections.

Project Structure

prompt_optimiser/
├── LICENSE.txt          # MIT license
├── index.js            # Main CLI application with vibe-lint rules
├── package-lock.json   # Dependency lock file  
└── package.json        # Package configuration and binary definition

Key Files

• index.js — Contains 20+ prompt linting rules, Claude API integration, and interactive CLI interface
• package.json — Defines the prompt-optimizer CLI command and project metadata

Contributing

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Add new rules to the RULES array in index.js following the existing pattern
4. Test your changes with various prompt inputs
5. Commit your changes (git commit -m 'Add amazing feature')
6. Push to the branch (git push origin feature/amazing-feature)
7. Open a Pull Request

Rule Structure

{
  id: "W010",
  severity: "warn", // or "error"  
  category: "category-name",
  pattern: /regex_pattern/gi,
  message: "Brief explanation of the problem",
  suggestions: [
    "Specific suggestion 1",
    "Specific suggestion 2"
  ]
}

License

MIT License - see LICENSE.txt for details.