Agent skill

github-analysis

Analyze GitHub commits, generate PR reviews, calculate contributor leaderboards, and assess code quality. Use when analyzing git commits, reviewing code, generating GitHub activity reports, or tracking developer contributions.

View SKILL.md on GitHub Repository
Stars 163
Forks 31

Install this agent skill to your Project

npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/data/github-analysis

SKILL.md

GitHub Analysis

Analyze GitHub activity, review code, and track contributions.

Quick Start

Analyze commits from JSON file:

bash
python scripts/analyze_commits.py commits.json

Generate leaderboard:

bash
python scripts/calculate_leaderboard.py commits.json --period week

Commit Analysis

What to Extract

From each commit, analyze:

  • Author & timestamp
  • Commit message quality
    • Clear (explains what and why)
    • Vague (just what, no why)
    • Cryptic (no context)
  • Files changed (count and types)
  • Lines added/removed
  • Code quality indicators
    • TODOs added
    • FIXMEs added
    • Console.log/debugging code
    • Commented code
    • Large file changes (>500 lines)

Quality Scoring

Commit Message Quality:

  • Excellent (8-10): Clear what + why, follows conventions
  • Good (5-7): Clear what, some context
  • Poor (1-4): Vague or no context
  • Bad (0): Single word, "wip", "test"

Code Quality Indicators:

bash
# Check for debugging code
grep -r "console.log\|debugger\|print(" changed_files/

# Check for TODOs
grep -r "TODO\|FIXME" changed_files/ | wc -l

# Check for commented code
grep -r "^[[:space:]]*//.*=\|^[[:space:]]*/\*" changed_files/

PR Review Template

Use this structure for code reviews:

markdown
# Pull Request Review

## Summary
[1-2 sentence overview of changes]

## Code Quality Assessment

### Structure & Organization
- ✅ **Good**: Well-organized, clear separation of concerns
- ⚠️  **Needs Work**: Mixed responsibilities, unclear structure
- 🔴 **Issues**: Significant structural problems

### Naming & Readability
- **Variables**: [clear/unclear/inconsistent]
- **Functions**: [descriptive/vague/confusing]
- **Comments**: [helpful/missing/outdated]

### Testing
- [ ] Unit tests included
- [ ] Integration tests updated
- [ ] Edge cases covered
- [ ] Test coverage: [%]

## Issues Found

### 🔴 Critical
- [Issue with security/correctness impact]

### 🟡 Warnings
- [Issue that should be addressed]

### 🔵 Suggestions
- [Nice-to-have improvements]

## Security Check

- [ ] No hardcoded credentials
- [ ] No SQL injection risks
- [ ] No XSS vulnerabilities
- [ ] Input validation present
- [ ] Authentication/authorization correct

## Performance

- [ ] No obvious performance issues
- [ ] Database queries optimized
- [ ] No N+1 query problems
- [ ] Appropriate caching

## Recommendations

1. [Priority recommendation]
2. [Additional improvement]
3. [Nice-to-have enhancement]

## Verdict

- [ ] ✅ **Approve** - Ready to merge
- [ ] 🟡 **Approve with Comments** - Minor issues, can merge
- [ ] 🔴 **Request Changes** - Must address issues before merge

Contributor Leaderboard

Metrics

Track these metrics per contributor:

  1. Commit Count (weight: 1x)
  2. Lines Changed (weight: 0.5x)
    • Added lines + modified lines
  3. Commit Quality (weight: 2x)
    • Average message quality score
  4. PR Reviews (weight: 1.5x)
    • Number of PR reviews contributed
  5. Response Time (weight: 1x)
    • Average time to respond to PR comments

Scoring Formula

Total Score = (commits × 1) +
              (lines_changed / 100 × 0.5) +
              (avg_quality × 2) +
              (pr_reviews × 1.5) +
              (10 - avg_response_hours × 1)

Leaderboard Format

markdown
# GitHub Contributor Leaderboard
## Period: [Week/Month]

| Rank | Contributor | Score | Commits | Lines | Quality | Reviews |
|------|-------------|-------|---------|-------|---------|---------|
| 1    | John Doe    | 45.2  | 12      | 1,234 | 8.5     | 5       |
| 2    | Jane Smith  | 38.7  | 10      | 987   | 7.8     | 4       |

Code Quality Metrics

Complexity Analysis

bash
# Count function complexity (rough estimate)
# Functions with >4 nested levels or >50 lines
grep -n "function\|def " file.js | while read line; do
    # Analyze complexity
done

Code Churn

Files with high churn (changed frequently):

bash
git log --format=format: --name-only | \
    sort | uniq -c | sort -rn | head -20

High churn may indicate:

  • Unstable code
  • Unclear requirements
  • Technical debt
  • Active development area

Test Coverage

bash
# Run test coverage (example)
npm test -- --coverage
python -m pytest --cov=src tests/

Good coverage targets:

  • Critical paths: 90%+
  • Business logic: 80%+
  • Overall: 70%+

Data Processing

Input Format (commits.json)

json
[
  {
    "sha": "abc123",
    "author": "John Doe",
    "email": "john@example.com",
    "date": "2026-01-04T10:30:00Z",
    "message": "Add user authentication feature",
    "files_changed": ["src/auth.js", "src/users.js"],
    "additions": 125,
    "deletions": 45,
    "files_count": 2
  }
]

Output Format (analysis.json)

json
{
  "summary": {
    "total_commits": 25,
    "total_contributors": 5,
    "total_files_changed": 67,
    "total_lines": 2345
  },
  "contributors": [
    {
      "name": "John Doe",
      "commits": 12,
      "lines_changed": 1234,
      "avg_quality": 8.5,
      "score": 45.2
    }
  ],
  "hot_files": [
    {"file": "src/auth.js", "changes": 8}
  ],
  "quality_issues": [
    {"type": "TODO", "count": 5},
    {"type": "console.log", "count": 3}
  ]
}

Scripts

analyze_commits.py

Analyzes commit data and generates metrics.

Usage:

bash
python scripts/analyze_commits.py input.json --output analysis.json

calculate_leaderboard.py

Calculates contributor rankings.

Usage:

bash
python scripts/calculate_leaderboard.py commits.json \
    --period week \
    --output leaderboard.json

generate_report.py

Generates HTML report from analysis.

Usage:

bash
python scripts/generate_report.py analysis.json \
    --template github-summary \
    --output report.html

Integration with Agents

Code Agent

python
# Get commits from GitHub
commits = github.get_commits(repo='owner/repo', days=7)

# Analyze with skill
python scripts/analyze_commits.py commits.json

Reporting Agent

python
# Generate leaderboard
python scripts/calculate_leaderboard.py commits.json

# Create HTML report
python scripts/generate_report.py analysis.json --template github-summary

Reference Files

  • metrics.md - Detailed scoring algorithms
  • patterns.md - Code quality patterns to detect
  • templates.md - Additional report templates

Recommended Agent Skills

Expand your agent's capabilities with these related and highly-rated skills.

majiayu000/claude-skill-registry

agent-ops-spec

Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.

163 31
Explore
majiayu000/claude-skill-registry

agent-ops-state

Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.

163 31
Explore
majiayu000/claude-skill-registry

agent-ops-spec

Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.

163 31
Explore
majiayu000/claude-skill-registry

agent-ops-testing

Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.

163 31
Explore
majiayu000/claude-skill-registry

agent-ops-testing

Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.

163 31
Explore
majiayu000/claude-skill-registry

agent-ops-state

Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.

163 31
Explore

Didn't find tool you were looking for?

Be as detailed as possible for better results