Debug Mining Engine

Debug Mining Engine

"The answer lies in the darkness" - Transform every debugging session into reusable skills

Overview

The Debug Mining Engine automatically captures debugging sessions and transforms them into three reusable assets:

1. Full Skills - Comprehensive methodologies with scripts and templates
2. Code Snippets - Quick copy/paste solutions
3. Pattern Guides - Conceptual understanding documents

Every bug you fix becomes a permanent asset in your skill arsenal.

When to Use This Skill

Use the Debug Mining Engine when you want to:

• Preserve debugging knowledge instead of losing it
• Build reusable solutions from errors you've fixed
• Create a growing library of debugging patterns
• Reduce time spent on similar issues in the future
• Transform failures into assets

The Philosophy

Traditional debugging:

• Fix the bug → Move on → Forget the solution
• Repeat the same debugging process later
• Knowledge lost, time wasted

Debug Mining:

• Fix the bug → Capture the session → Generate skills
• Reuse the solution next time
• Knowledge preserved, time saved

ROI: Every debugging session becomes an investment that pays dividends forever.

How It Works

Hybrid Detection System

AI Auto-Detection (Passive):

• Monitors shell for error patterns
• Detects failures automatically
• Starts background capture
• Prompts you after detecting a fix

Manual Override (Active):

/debug-start "Description of what you're debugging"
# ... debug and fix the issue ...
/debug-end

Smart Prompts:

🤔 I noticed you just fixed an error. Should I save this debugging session?
   Error: "Could not find table 'repositories'"
   Solution: Created schema validation script
   
   [Yes, save it] [No, discard] [Let me review first]

What Gets Captured

• Error Context: Full error message, stack trace, exit code
• Solution Journey: All attempts (including failures)
• Final Solution: The code/command that worked
• Environment: Working directory, relevant files
• Metadata: Duration, timestamps, tags

What Gets Generated

1. Full Skill (/home/ubuntu/skills/skills/debugging-patterns/[skill-name]/)

├── SKILL.md                    # Complete methodology
├── scripts/
│   ├── detect_error.py         # Error detection
│   ├── apply_fix.py            # Automated fix
│   └── validate_solution.py    # Validation
├── templates/
│   ├── fix_template.sh         # Template for similar fixes
│   └── test_template.py        # Test template
└── references/
    ├── error_analysis.md       # Deep dive
    ├── solution_rationale.md   # Why this works
    └── related_patterns.md     # Similar issues

2. Code Snippet (/home/ubuntu/debug-snippets/[category]/[name].sh)

#!/bin/bash
# Quick Fix: Supabase Schema Validation
# Error: Could not find table 'repositories'
# Solution: Validate schema before querying

python3 << 'EOF'
from supabase import create_client
client = create_client(url, key)

# Validate table exists
try:
    client.table('your_table').select('*').limit(1).execute()
    print("✅ Table exists")
except Exception as e:
    print(f"❌ Table not found: {e}")
EOF

3. Pattern Guide (/home/ubuntu/debug-patterns/[name].md)

# Pattern: Supabase Schema Validation

## The Problem
Querying Supabase tables without validating schema first

## Why This Happens
- Table names change during development
- Schema migrations not applied

## The Solution Pattern
Always validate schema before querying

## When to Apply
- Before any Supabase query
- After schema migrations

Setup

Step 1: Install the Monitoring System

# Run the setup script
bash /home/ubuntu/skills/skills/utility-skills/debug-mining-engine/scripts/setup_monitor.sh

This adds debug monitoring to your shell environment.

Step 2: Create Storage Directories

mkdir -p /home/ubuntu/debug-sessions
mkdir -p /home/ubuntu/debug-snippets/{database,api,filesystem,network,auth}
mkdir -p /home/ubuntu/debug-patterns

Step 3: Test the System

# Trigger a test error
/debug-start "Testing debug mining"
python3 -c "raise Exception('Test error')"
/debug-end

# Check if session was captured
ls /home/ubuntu/debug-sessions/$(date +%Y-%m-%d)/

Usage Examples

Example 1: Automatic Capture

# You run a command that fails
$ python3 audit_script.py
❌ Error: Could not find table 'repositories'

# AI detects error and starts capturing
[Debug Mining: Capture started]

# You try different solutions
$ python3 audit_script.py --table users
❌ Still failing

$ python3 audit_script.py --validate-schema
✅ Success!

# AI detects fix and prompts
🤔 Should I save this debugging session?
[Yes, save it]

# AI generates assets
✅ Generated 3 assets:
   1. Full Skill: supabase-schema-validator
   2. Code Snippet: supabase-schema-check.sh
   3. Pattern Guide: database-validation.md

Example 2: Manual Capture

# Start debugging manually
$ /debug-start "Fixing API connection timeout"
[Debug Mining: Manual capture started]

# Debug and fix
$ curl https://api.example.com/endpoint
Error: Connection timeout

$ curl --retry 3 --retry-delay 2 https://api.example.com/endpoint
✅ Success!

# End capture
$ /debug-end
[Debug Mining: Analyzing session...]

✅ Generated skill: api-retry-with-backoff

Example 3: Using Generated Skills

# List available debugging skills
ls /home/ubuntu/skills/skills/debugging-patterns/

# Use a code snippet
bash /home/ubuntu/debug-snippets/database/supabase-schema-check.sh

# Read a pattern guide
cat /home/ubuntu/debug-patterns/database-validation.md

Intelligence Features

Pattern Recognition

The system learns from your debugging history:

🔍 Pattern Detected!
   You've debugged "table not found" errors 3 times this month.
   
   Suggestion: Create a pre-flight validation skill
   
   [Create skill] [Remind me later]

Proactive Suggestions

💡 Based on your debugging history:
   - "api-connection-validator" (you debug API errors often)
   - "environment-config-checker" (env vars cause 40% of errors)
   
   [Generate these skills] [Show examples]

Learning Metrics

📈 Debug Mining Stats (Last 30 Days)
   
   Debugging Time: 12 hours → 3 hours (-75%)
   Errors Captured: 45
   Skills Generated: 12
   Reuse Count: 28
   Time Saved: ~9 hours

Scripts Reference

Core Scripts

setup_monitor.sh - Install shell monitoring system

bash scripts/setup_monitor.sh

monitor.py - Error detection and session capture

python3 scripts/monitor.py --exit-code 1 --command "failed_command"

analyzer.py - Pattern analysis and extraction

python3 scripts/analyzer.py --session-id debug_2026-02-11_01-33-08

generator.py - Multi-format skill generation

python3 scripts/generator.py --session-id debug_2026-02-11_01-33-08

Utility Scripts

list_sessions.py - List captured debugging sessions

python3 scripts/list_sessions.py --days 30

stats.py - Show debug mining statistics

python3 scripts/stats.py

search_patterns.py - Search for similar patterns

python3 scripts/search_patterns.py --error "table not found"

Templates Reference

skill_template/ - Template for generated full skills snippet_template.sh - Template for code snippets pattern_template.md - Template for pattern guides

Best Practices

1. Add Context When Starting Manual Capture

# Good
/debug-start "Fixing Supabase query error in audit script"

# Not as good
/debug-start "debugging"

2. Review Generated Skills

Always review generated skills before adding to arsenal:

# Review before committing
cat /home/ubuntu/skills/skills/debugging-patterns/new-skill/SKILL.md

3. Tag Your Sessions

Add tags to make skills discoverable:

/debug-start "Fixing API timeout" --tags api,network,timeout

4. Update Existing Skills

If you find a better solution, update the skill:

python3 scripts/update_skill.py --skill supabase-schema-validator --session debug_2026-02-12_10-30-00

5. Share Patterns

Export patterns to share with team:

python3 scripts/export_pattern.py --pattern database-validation --format markdown

Troubleshooting

Monitor Not Detecting Errors

Problem: Errors not being captured automatically

Solution:

1. Check if monitor is installed: which debug_monitor_command
2. Reload shell: source ~/.bashrc
3. Test manually: /debug-start "test"

Sessions Not Saving

Problem: Debug sessions not appearing in /home/ubuntu/debug-sessions/

Solution:

1. Check directory permissions: ls -la /home/ubuntu/debug-sessions/
2. Check monitor logs: tail -f /home/ubuntu/debug-mining/monitor.log
3. Run monitor manually: python3 scripts/monitor.py --test

Generated Skills Have Errors

Problem: Generated skills contain incorrect code

Solution:

1. Review the captured session: cat /home/ubuntu/debug-sessions/[date]/[session-id].json
2. Regenerate with corrections: python3 scripts/generator.py --session [id] --review
3. Edit manually and mark as reviewed

Success Metrics

Capture Quality

• Accuracy: 95%+ of debugging sessions correctly detected
• Completeness: 90%+ of sessions with full context
• False Positives: <5% non-debugging sessions captured

Skill Quality

• Reusability: Generated skills used 3+ times on average
• Time Saved: 2-5 hours saved per reused skill
• Pattern Coverage: 80%+ of errors have matching patterns

System Impact

• Debugging Time Reduction: 50-75% decrease
• Knowledge Retention: 100% of debugging knowledge preserved
• Arsenal Growth: 10-20 new skills per month

Advanced Features

Custom Analyzers

Create custom pattern analyzers for your specific domain:

# /home/ubuntu/debug-mining/custom_analyzers/my_analyzer.py

from analyzer import PatternAnalyzer

class MyCustomAnalyzer(PatternAnalyzer):
    def analyze_custom_pattern(self, session):
        # Your custom analysis logic
        pass

Integration with Skill Arsenal

Generated skills automatically integrate with your unified skill arsenal:

from lib.skill_registry import SkillRegistry

registry = SkillRegistry('/home/ubuntu/skills/skills.json')

# Find debugging skills
debug_skills = registry.find_by_category('debugging-patterns')

# Get related skills
related = registry.get_complements('supabase-schema-validator')

Workflow Automation

Chain debugging skills with other skills:

from lib.skill_composer import SkillComposer

composer = SkillComposer(registry)

# Create workflow
workflow = composer.compose_workflow([
    'brainstorming',
    'writing-plans',
    'database-schema-generator',
    'supabase-schema-validator',  # Generated from debugging!
    'testing-framework'
])

composer.execute_workflow(workflow, context)

Future Enhancements

• AI-powered root cause analysis using LLMs
• Cross-project pattern recognition
• Team collaboration (share debugging patterns)
• Visual debugging timeline
• Integration with systematic-debugging skill

Related Skills

• systematic-debugging - Methodical debugging approach
• skill-arsenal-builder - Build unified skill collections
• skill-creator - Create new skills from scratch
• feature-verification - Verify fixes work correctly

Time Investment

• Setup: 15 minutes (one-time)
• Per debugging session: 0 minutes (automatic)
• Reviewing generated skills: 5-10 minutes
• ROI: 50-75% reduction in debugging time

Key Takeaway

Every bug you fix becomes a permanent asset.

Stop losing debugging knowledge. Start building a library of solutions that grows with every error you encounter.

"The answer lies in the darkness" - And now you can capture it! 🎯