issue-duplicate-detection

# GitHub Issue Duplicate Detection Skill

This skill provides patterns for detecting and handling duplicate GitHub issues.

## Why Check for Duplicates

- Prevent duplicate work
- Consolidate discussion
- Better tracking and metrics
- Cleaner backlog

---

## Detection Process

### Step 1: Extract Key Terms

From the potential new issue:
- Core functionality/feature
- Error messages (exact text)
- Component names
- User actions described

### Step 2: Search Existing Issues

```bash
# Search open issues
gh issue list --state open --search "keyword1 keyword2" --limit 20

# Search all issues (including closed)
gh issue list --state all --search "keyword1 keyword2" --limit 20

# Search with specific terms
gh issue list --state all --search "in:title error message"

# Search by label
gh issue list --state open --label "bug" --search "keyword"
```

### Step 3: Review Results

For each potential match:
- Compare titles carefully
- Read descriptions
- Check if closed issue was fixed
- Consider different wordings of same problem

### Step 4: Present Findings

```markdown
## Duplicate Check Results

### Potential Duplicates Found

| # | Title | State | Created | Similarity |
|---|-------|-------|---------|------------|
| #123 | Similar issue title | Open | 2024-01-10 | High |
| #456 | Related problem | Closed | 2023-12-01 | Medium |

### Analysis

**#123** appears to describe the same issue:
- Same error message mentioned
- Same component affected
- Both triggered by same user action

**Recommendation**: Add context to #123 instead of creating new issue.

### If Not Duplicate

If you confirm this is a new issue, I'll proceed with creation.
```

---

## Search Strategies

### By Error Message
```bash
gh issue list --state all --search "TypeError: Cannot read property"
```

### By Component/Feature
```bash
gh issue list --state all --search "in:title checkout"
gh issue list --state all --search "in:body payment"
```

### By Author (Recent)
```bash
gh issue list --state all --au