Debug

/debug

Usage

/debug $ARGUMENTS              # Debug the issue
/debug --performance <issue>            # Focus on performance debugging
/debug --issue <issue_description>      # Debug and create GitHub issue

Description

Systematic debugging using a single debugger agent. Investigates root causes, proposes fixes, and optionally creates GitHub issues from findings.

Behavior

Uses explicit task phases for progress visibility:

Task Phases:
  1. "Investigate root cause"
     - Deploy debugger agent
     - Gather evidence and logs
     - Identify probable cause
     - Status: pending → in_progress → completed

  2. "Implement fix"
     - Apply targeted fix based on findings
     - Minimal changes to address root cause
     - Status: pending → in_progress → completed

  3. "Verify fix"
     - Run relevant tests
     - Confirm issue resolved
     - Check for regressions
     - Status: pending → in_progress → completed

Execution Steps

1. Analyze: Deploy debugger agent to investigate the issue
2. Fix: Implement the fix based on findings
3. Verify: Run tests to confirm the fix works

With --issue Flag

After debugging, create a GitHub issue documenting:

• Root cause analysis
• Steps to reproduce
• Fix applied (if any)
• Prevention recommendations

Expected Output

User: /debug Memory leak causing crashes

Investigating: Memory leak causing crashes

Deploying debugger agent...

Root Cause Analysis:
  Issue: Event listeners not cleaned up in useEffect
  Evidence: Heap growing 50MB/hour, correlates with component mounts
  Confidence: 92%

Fix Applied:
  - Added cleanup function to useEffect in UserSession.tsx
  - Implemented proper listener removal

Verification:
  - Tests passing
  - Memory stable after 10 minute test

Issue resolved

With --issue Flag

User: /debug --issue Payment timeout

[Debug analysis completes...]

Creating GitHub issue...
Issue #142 created: "Payment timeout - Database connection pool exhaustion"

Notes

• Single debugger agent handles all investigation
• Use --performance for CPU/memory/latency issues
• Use --issue to document findings as GitHub issue
• Typical execution: 2-5 minutes