Initial commit

skills/skill-customizer/scripts/track_feedback.py

@@ -0,0 +1,257 @@
+#!/usr/bin/env python3
+"""
+Feedback Tracker - Captures and organizes user feedback for skill customization
+
+This script helps structure user feedback about skill performance, making it
+easier to identify specific areas for improvement and track customization needs.
+
+Usage:
+    track_feedback.py <skill-directory>
+
+Interactive mode:
+    The script will guide you through collecting structured feedback about:
+    - What task was attempted
+    - What worked well
+    - What didn't match expectations
+    - Specific preferences or requirements
+    - Suggested improvements
+
+Output:
+    Creates/appends to FEEDBACK.md in the skill directory with structured feedback entries
+"""
+
+import sys
+from pathlib import Path
+from datetime import datetime
+
+
+FEEDBACK_TEMPLATE = """# Skill Feedback Log
+
+This file tracks user feedback and customization needs for iterative skill improvement.
+
+---
+"""
+
+
+FEEDBACK_ENTRY_TEMPLATE = """
+## Feedback Entry #{entry_num} - {date}
+
+### Task Context
+**What were you trying to accomplish?**
+{task_description}
+
+### What Worked Well
+{what_worked}
+
+### What Didn't Match Expectations
+{what_didnt_work}
+
+### Specific Preferences/Requirements
+{preferences}
+
+### Suggested Improvements
+{improvements}
+
+### Priority
+{priority}
+
+---
+"""
+
+
+def collect_feedback_interactive():
+    """
+    Interactively collect structured feedback from user.
+
+    Returns:
+        Dictionary with feedback components
+    """
+    print("\n📝 Skill Feedback Collection")
+    print("=" * 60)
+    print("Please provide detailed feedback to help customize this skill.\n")
+
+    feedback = {}
+
+    # Task context
+    print("1️⃣  What task were you trying to accomplish?")
+    print("   (Be specific: e.g., 'Extract tables from a 20-page PDF report')")
+    feedback['task_description'] = input("   > ").strip()
+    print()
+
+    # What worked
+    print("2️⃣  What aspects of the skill worked well?")
+    print("   (e.g., 'Text extraction was accurate', 'Fast processing')")
+    feedback['what_worked'] = input("   > ").strip()
+    if not feedback['what_worked']:
+        feedback['what_worked'] = "N/A"
+    print()
+
+    # What didn't work
+    print("3️⃣  What didn't match your expectations?")
+    print("   (e.g., 'Output format was JSON, I needed Markdown', 'Too verbose')")
+    feedback['what_didnt_work'] = input("   > ").strip()
+    if not feedback['what_didnt_work']:
+        feedback['what_didnt_work'] = "N/A"
+    print()
+
+    # Preferences
+    print("4️⃣  What are your specific preferences or requirements?")
+    print("   (e.g., 'Always output in Markdown', 'Include source page numbers')")
+    feedback['preferences'] = input("   > ").strip()
+    if not feedback['preferences']:
+        feedback['preferences'] = "N/A"
+    print()
+
+    # Improvements
+    print("5️⃣  What specific improvements would help?")
+    print("   (e.g., 'Add option to filter by date range', 'Default to concise output')")
+    feedback['improvements'] = input("   > ").strip()
+    if not feedback['improvements']:
+        feedback['improvements'] = "N/A"
+    print()
+
+    # Priority
+    print("6️⃣  How important is this customization?")
+    print("   Options: critical, high, medium, low")
+    priority = input("   > ").strip().lower()
+    if priority not in ['critical', 'high', 'medium', 'low']:
+        priority = 'medium'
+    feedback['priority'] = priority.capitalize()
+    print()
+
+    return feedback
+
+
+def save_feedback(skill_dir, feedback):
+    """
+    Save feedback to FEEDBACK.md file.
+
+    Args:
+        skill_dir: Path to skill directory
+        feedback: Dictionary with feedback components
+
+    Returns:
+        Path to feedback file
+    """
+    feedback_path = skill_dir / 'FEEDBACK.md'
+
+    # Create feedback file if it doesn't exist
+    if not feedback_path.exists():
+        feedback_path.write_text(FEEDBACK_TEMPLATE)
+        entry_num = 1
+    else:
+        # Count existing entries
+        content = feedback_path.read_text()
+        entry_num = content.count('## Feedback Entry #') + 1
+
+    # Format feedback entry
+    entry = FEEDBACK_ENTRY_TEMPLATE.format(
+        entry_num=entry_num,
+        date=datetime.now().strftime("%Y-%m-%d %H:%M"),
+        task_description=feedback['task_description'],
+        what_worked=feedback['what_worked'],
+        what_didnt_work=feedback['what_didnt_work'],
+        preferences=feedback['preferences'],
+        improvements=feedback['improvements'],
+        priority=feedback['priority']
+    )
+
+    # Append to file
+    with open(feedback_path, 'a') as f:
+        f.write(entry)
+
+    return feedback_path
+
+
+def analyze_feedback_patterns(feedback_path):
+    """
+    Analyze feedback file for common patterns.
+
+    Args:
+        feedback_path: Path to feedback file
+
+    Returns:
+        Dictionary with analysis results
+    """
+    if not feedback_path.exists():
+        return None
+
+    content = feedback_path.read_text()
+
+    analysis = {
+        'total_entries': content.count('## Feedback Entry #'),
+        'critical_priority': content.count('Priority\nCritical'),
+        'high_priority': content.count('Priority\nHigh'),
+        'common_themes': []
+    }
+
+    # Simple keyword analysis for common themes
+    keywords = {
+        'output format': ['format', 'markdown', 'json', 'output'],
+        'verbosity': ['verbose', 'concise', 'too much', 'too little'],
+        'performance': ['slow', 'fast', 'speed', 'performance'],
+        'accuracy': ['accurate', 'wrong', 'incorrect', 'missing'],
+    }
+
+    for theme, words in keywords.items():
+        if any(word.lower() in content.lower() for word in words):
+            analysis['common_themes'].append(theme)
+
+    return analysis
+
+
+def main():
+    if len(sys.argv) < 2:
+        print("Usage: track_feedback.py <skill-directory>")
+        print("\nThis script helps collect structured feedback for skill customization.")
+        print("Run it after using a skill to capture what needs to be improved.")
+        sys.exit(1)
+
+    skill_path = Path(sys.argv[1]).resolve()
+
+    # Validate skill directory
+    if not skill_path.exists():
+        print(f"❌ Error: Skill directory not found: {skill_path}")
+        sys.exit(1)
+
+    if not skill_path.is_dir():
+        print(f"❌ Error: Path is not a directory: {skill_path}")
+        sys.exit(1)
+
+    skill_md = skill_path / 'SKILL.md'
+    if not skill_md.exists():
+        print(f"❌ Error: Not a valid skill directory (SKILL.md not found): {skill_path}")
+        sys.exit(1)
+
+    print(f"📂 Skill: {skill_path.name}")
+
+    # Collect feedback
+    feedback = collect_feedback_interactive()
+
+    # Save feedback
+    try:
+        feedback_path = save_feedback(skill_path, feedback)
+        print(f"\n✅ Feedback saved to: {feedback_path.name}")
+    except Exception as e:
+        print(f"\n❌ Error saving feedback: {e}")
+        sys.exit(1)
+
+    # Analyze patterns
+    analysis = analyze_feedback_patterns(feedback_path)
+    if analysis and analysis['total_entries'] > 1:
+        print(f"\n📊 Feedback Summary:")
+        print(f"   Total entries: {analysis['total_entries']}")
+        print(f"   Critical priority: {analysis['critical_priority']}")
+        print(f"   High priority: {analysis['high_priority']}")
+        if analysis['common_themes']:
+            print(f"   Common themes: {', '.join(analysis['common_themes'])}")
+
+    print(f"\n💡 Next steps:")
+    print(f"   1. Review feedback in {feedback_path.name}")
+    print(f"   2. Identify specific changes to make in SKILL.md or scripts")
+    print(f"   3. Apply improvements and test")
+    print(f"   4. Document changes in CUSTOMIZATION_LOG.md")
+
+
+if __name__ == "__main__":
+    main()