Bug Hunter

Analyze, reproduce, and report bugs with precision.

When to Use

✅ USE this skill when:

• Analyzing error logs, stack traces, crash dumps
• Writing structured bug reports
• Triaging and prioritizing bugs
• Root cause analysis from symptoms
• "这个报错是什么原因" / "帮我写个bug单"

❌ DON'T use this skill when:

• Fixing the code → use coding tools
• Writing test cases → use test-case-gen
• Testing APIs → use api-tester

Bug Report Template

When writing a bug report, use this structure:

## 🐛 Bug Report

**Title**: [Module] Brief description of the issue

**Severity**: 🔴 Critical / 🟠 Major / 🟡 Minor / 🟢 Trivial
**Priority**: P0-Blocker / P1-High / P2-Medium / P3-Low
**Environment**: OS / Browser / App Version / API Version
**Reporter**: 虫探 🔍
**Date**: YYYY-MM-DD

### Description
Clear, concise description of what went wrong.

### Steps to Reproduce
1. Step one (be specific: URL, button name, exact input)
2. Step two
3. Step three

### Test Data
- Account: xxx
- Input: xxx

### Expected Result
What should happen.

### Actual Result
What actually happened. Include error messages verbatim.

### Evidence
- Screenshot: [attached]
- Log snippet:
\`\`\`
ERROR 2024-01-01 10:00:00 NullPointerException at UserService.java:42
\`\`\`

### Root Cause Analysis (if identified)
- Location: file:line
- Cause: description
- Impact scope: what else might be affected

### Suggested Fix (if obvious)
Brief suggestion for the developer.

Log Analysis

Common Error Patterns

Log Analysis Steps

1. Identify the error: Find the first error in the chain (root cause, not symptom)
2. Check timestamp: When did it first occur? Is it recurring?
3. Check context: What request/operation triggered it?
4. Check stack trace: Which module/function? What line?
5. Check related logs: What happened before the error?
6. Reproduce: Can you trigger the same error consistently?

# Quick log analysis commands
# Find errors in log file
grep -n -i "error\|exception\|fatal\|failed" app.log | tail -20

# Count error types
grep -oP '\w+Exception' app.log | sort | uniq -c | sort -rn

# Find errors in time range
awk '/2024-01-01 10:0[0-5]/' app.log | grep -i error

# Get context around an error (5 lines before and after)
grep -B5 -A5 "NullPointerException" app.log

Severity Classification

Triage Decision Matrix

                    High Impact          Low Impact
High Frequency   →  P0 Fix Now          P1 Fix Soon
Low Frequency    →  P1 Fix Soon         P2/P3 Backlog

Root Cause Categories

When analyzing root cause, classify into:

• Code Defect: Logic error, missing validation, wrong algorithm
• Config Error: Wrong environment config, missing feature flag
• Data Issue: Corrupt data, migration problem, encoding issue
• Infrastructure: Server capacity, network, third-party service
• Design Flaw: Architectural issue, race condition by design
• Requirement Gap: Ambiguous or missing requirement

Tips

• Always include the exact error message, not a paraphrase
• Screenshots > descriptions for UI bugs
• Include both expected AND actual results — never skip either
• Note if the bug is intermittent (include frequency)
• Check if the bug exists in previous versions (regression?)
• For API bugs, include the full request and response

Screenshot / UI Bug Analysis

When provided with screenshots:

1. Describe the visual issue: Element misaligned? Missing? Wrong state?
2. Identify affected component: Which UI component/area?
3. Check viewport: Is it responsive-related? Note screen size if visible.
4. Compare: Expected design vs actual rendering
5. Classify: Layout bug / Style bug / Functional state bug / Data display bug

Use the image tool to analyze screenshots, then include findings in the bug report's Evidence section.

API Error Response Analysis

When analyzing API errors:

# Save full request + response for bug report
curl -v -X POST "https://api.example.com/endpoint" \
  -H "Content-Type: application/json" \
  -d '{"key": "value"}' 2>&1 | tee /tmp/api-debug.log

# Key things to capture:
# - Request URL, method, headers, body
# - Response status code, headers, body
# - Timing information
# - Any redirect chain

Common API error patterns: