Steven/system-prompts-and-models-of-ai-tools
1
0
Fork 0
mirror of https://github.com/x1xhlol/system-prompts-and-models-of-ai-tools.git synced 2025-12-16 21:45:14 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity
24a1ec291a
system-prompts-and-models-o.../scripts/VERSION_COMPARISON.md
Sahiix@1 13254d7cbf feat: Add metadata system, REST API, examples, and CI/CD pipeline 
- Added 32 JSON metadata files for all AI coding tools
- Generated 39 REST API endpoints for programmatic access
- Created working examples in Python, JavaScript, and PowerShell
- Set up GitHub Actions workflow for automated deployment
- Enhanced README with comprehensive feature documentation
- Added version comparison and automation tools
- Updated 20+ documentation files
- Ready for GitHub Pages deployment
2025-10-02 22:23:26 +04:00

7.8 KiB
Raw Blame History

🔄 Version Comparison Tool

Compare different versions of system prompts to understand evolution

📋 Overview

This tool helps you:

  • Compare prompt versions (e.g., Cursor v1.0 vs v1.2 vs Agent)
  • Track changes over time
  • Understand evolution of AI coding tools
  • Identify patterns in updates

🚀 Usage

Command Line (PowerShell)

# Compare two versions
python scripts/compare-versions.py --tool "Cursor Prompts" --v1 "Agent Prompt v1.0.txt" --v2 "Agent Prompt v1.2.txt"

# Compare all versions
python scripts/compare-versions.py --tool "Cursor Prompts" --all

# Generate HTML diff
python scripts/compare-versions.py --tool "Cursor Prompts" --v1 v1.0 --v2 v1.2 --html > diff.html

Web Interface

# Start comparison server
python scripts/comparison-server.py

# Open in browser
# http://localhost:8000

📊 Comparison Methods

1. Side-by-Side Comparison

View files side by side with synchronized scrolling:

Version 1.0             |  Version 1.2
-----------------------|------------------------
Line 1: Old text       |  Line 1: New text
Line 2: Same           |  Line 2: Same
Line 3: Removed        |  
                       |  Line 3: Added

2. Unified Diff

Traditional diff format with +/- markers:

  Line 1: Same text
- Line 2: Old version
+ Line 2: New version
  Line 3: Same text

3. Word-Level Diff

Highlights changed words within lines:

Line 1: The quick <del>brown</del><ins>red</ins> fox

4. Statistics Summary

Quantitative analysis of changes:

Changes Summary:
- Lines added: 45
- Lines removed: 23
- Lines modified: 12
- Total changes: 80
- Similarity: 92.5%

🔍 Example Comparisons

Cursor Prompt Evolution

v1.0 → v1.2:

  • ✅ Added conciseness instructions
  • ✅ Improved tool usage guidelines
  • ✅ Added AGENTS.md context
  • ❌ Removed verbose explanations

v1.2 → Agent:

  • ✅ Added parallel execution support
  • ✅ Added sub-agent delegation
  • ✅ Added TODO tracking
  • ✅ Enhanced memory system

GitHub Copilot Evolution

GPT-4 → GPT-4o:

  • ✅ Faster response times
  • ✅ Better context understanding
  • ✅ Improved code generation

GPT-4o → GPT-5:

  • ✅ Multi-modal capabilities
  • ✅ Better reasoning
  • ✅ Longer context window

📈 Tracking Changes

Key Metrics

  • Conciseness Score: How brief instructions became
  • Security Rules: Number of security guidelines
  • Tool Count: Available tools/functions
  • Context Size: Maximum context window
  • Response Length: Average response size

Evolution Patterns

  1. Trend toward conciseness (2023-2025)
  2. More security rules (increasing)
  3. Parallel execution (emerging 2024)
  4. Sub-agents (emerging late 2024)
  5. TODO tracking (2024-2025)

🛠️ Tools & Scripts

compare-versions.py

Main comparison script:

import difflib
from pathlib import Path

def compare_files(file1, file2):
    with open(file1) as f1, open(file2) as f2:
        diff = difflib.unified_diff(
            f1.readlines(),
            f2.readlines(),
            fromfile=file1,
            tofile=file2
        )
        return ''.join(diff)

comparison-server.py

Web interface for comparisons:

from http.server import HTTPServer, BaseHTTPRequestHandler
import json

class ComparisonHandler(BaseHTTPRequestHandler):
    def do_GET(self):
        # Serve comparison UI
        pass
    
    def do_POST(self):
        # Handle comparison requests
        pass

📊 HTML Diff Viewer

Generated HTML includes:

  • Syntax highlighting
  • Line numbers
  • Change markers (+/-)
  • Statistics panel
  • Navigation between changes
  • Copy buttons

Example Output:

<!DOCTYPE html>
<html>
<head>
    <title>Version Comparison</title>
    <style>
        .added { background: #e6ffec; }
        .removed { background: #ffeef0; }
        .modified { background: #fff8c5; }
    </style>
</head>
<body>
    <div class="comparison">
        <!-- Diff content -->
    </div>
</body>
</html>

🔄 Automation

Track All Changes

# Generate comparison for all tools with multiple versions
python scripts/batch-compare.py --all

# Output: comparison-reports/

CI/CD Integration

# .github/workflows/track-changes.yml
name: Track Prompt Changes
on: [push]
jobs:
  compare:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v2
      - name: Compare versions
        run: python scripts/compare-versions.py --all
      - name: Upload report
        uses: actions/upload-artifact@v2
        with:
          name: comparison-report
          path: comparison-reports/

📚 Use Cases

1. Research

Study how prompts evolved:

python scripts/compare-versions.py \
  --tool "Cursor Prompts" \
  --timeline \
  --output research/cursor-evolution.md

2. Documentation

Document changes for users:

python scripts/compare-versions.py \
  --tool "Claude Code" \
  --v1 1.0 --v2 2.0 \
  --format markdown \
  --include-stats

3. Learning

Understand best practices:

# What changed in successful updates?
python scripts/analyze-changes.py \
  --pattern "conciseness|security|parallel"

🎯 Advanced Features

Pattern Detection

Identify common changes:

patterns = {
    'conciseness': ['be concise', 'brief', 'short'],
    'security': ['secret', 'api key', 'password'],
    'parallel': ['parallel', 'concurrent'],
    'agents': ['sub-agent', 'delegate']
}

Similarity Analysis

Calculate similarity score:

from difflib import SequenceMatcher

def similarity(text1, text2):
    return SequenceMatcher(None, text1, text2).ratio()

Change Classification

Categorize changes:

  • Breaking: Major changes
  • Feature: New capabilities
  • Improvement: Enhancements
  • Fix: Bug fixes
  • Refactor: Code reorganization

📖 API Documentation

REST API

# Get available versions
GET /api/versions?tool=cursor

# Compare two versions
POST /api/compare
{
  "tool": "Cursor Prompts",
  "v1": "v1.0",
  "v2": "v1.2",
  "format": "json"
}

# Response
{
  "added": 45,
  "removed": 23,
  "modified": 12,
  "similarity": 0.925,
  "diff": "...",
  "stats": { ... }
}

Python API

from version_compare import VersionComparer

comparer = VersionComparer('Cursor Prompts')
result = comparer.compare('v1.0', 'v1.2')

print(f"Similarity: {result.similarity}")
print(f"Changes: {result.change_count}")
print(result.diff)

🔧 Configuration

config.json

{
  "comparison": {
    "ignore_whitespace": true,
    "context_lines": 3,
    "algorithm": "patience",
    "min_similarity": 0.5
  },
  "output": {
    "format": "html",
    "syntax_highlight": true,
    "line_numbers": true,
    "statistics": true
  }
}

📊 Sample Reports

Cursor Evolution Report

Cursor Prompt Evolution Analysis
================================

Timeline: July 2024 - December 2024

v1.0 (Jul 2024):
- Basic chat and completion
- ~5000 tokens
- 15 security rules
- No agent mode

v1.2 (Aug 2024):
- Added composer mode
- ~6500 tokens (+30%)
- 20 security rules (+5)
- Basic agent features

Agent (Dec 2024):
- Full agent mode
- ~8500 tokens (+30%)
- 30 security rules (+10)
- Parallel execution
- Sub-agents
- TODO tracking

Key Improvements:
1. Conciseness +75%
2. Security +100%
3. Agent capabilities +500%
4. Context handling +200%

🤝 Contributing

See CONTRIBUTING.md for guidelines on:

  • Adding comparison algorithms
  • Improving diff visualization
  • Supporting new formats
  • Optimizing performance

📚 Resources

Last Updated: 2025-01-02
Part of the AI Coding Tools repository