GitHub Repository Analysis Skill

GitHub Repository Analysis

This skill guides analysis of GitHub repositories to extract meaningful insights about development activity, contributor patterns, and release cycles.

Core Analysis Capabilities

Commit Frequency Analysis

Extract commit history via GitHub API or git commands
Calculate commit frequency by time period (daily, weekly, monthly)
Identify patterns in development activity
Detect periods of high/low activity
Generate time-series visualizations of commit trends

Outstanding Contributors Analysis

Identify top contributors by commit count
Calculate contribution distribution (percentage per contributor)
Analyze commit patterns by contributor
Track first-time vs. recurring contributors
Generate contributor leaderboards

Release Timeline Analysis

Extract release/tag history
Calculate time between releases
Identify release patterns and cycles
Track version numbering schemes
Map releases to major commit periods

Implementation Approaches

Approach 1: GitHub API (Recommended for Public Repos)

Use GitHub's REST or GraphQL API for efficient data retrieval:

import requests
from datetime import datetime

def analyze_commits(owner, repo, token=None):
    headers = {'Authorization': f'token {token}'} if token else {}
    url = f'https://api.github.com/repos/{owner}/{repo}/commits'
    
    all_commits = []
    page = 1
    
    while True:
        response = requests.get(url, headers=headers, params={'page': page, 'per_page': 100})
        commits = response.json()
        if not commits:
            break
        all_commits.extend(commits)
        page += 1
    
    return all_commits

def analyze_contributors(commits):
    contributor_stats = {}
    for commit in commits:
        author = commit['commit']['author']['name']
        contributor_stats[author] = contributor_stats.get(author, 0) + 1
    
    return sorted(contributor_stats.items(), key=lambda x: x[1], reverse=True)

def analyze_releases(owner, repo, token=None):
    headers = {'Authorization': f'token {token}'} if token else {}
    url = f'https://api.github.com/repos/{owner}/{repo}/releases'
    response = requests.get(url, headers=headers)
    return response.json()

Benefits:

No repository cloning needed
Efficient pagination
Access to additional metadata
Rate limiting: 60 requests/hour without auth, 5000 with token

Approach 2: Local Git Repository

Use git commands for detailed analysis when repository is already cloned:

# Get commit history with timestamps
git log --pretty=format:"%h|%an|%ae|%ad|%s" --date=iso > commits.txt

# Count commits by author
git shortlog -sn --all

# Get all tags/releases
git tag -l --sort=-version:refname

# Commit frequency by week
git log --pretty=format:"%ad" --date=short | awk '{print $1}' | uniq -c

# Commits per month
git log --pretty=format:"%ad" --date=format:"%Y-%m" | sort | uniq -c

Approach 3: Hybrid Approach

Combine both methods for comprehensive analysis:

Use API for releases and high-level stats
Clone repository for detailed commit analysis
Use git commands for advanced filtering

Analysis Workflow

Repository Identification
- Parse GitHub URL or accept owner/repo parameters
- Validate repository exists and is accessible
Data Collection
- Fetch commit history (API or git log)
- Retrieve release/tag information
- Collect contributor metadata
Data Processing
- Parse timestamps and author information
- Group commits by time periods
- Calculate statistics and metrics
Insight Generation
- Identify top contributors with percentages
- Calculate commit frequency trends
- Map release timeline with intervals
- Detect anomalies or interesting patterns
Visualization & Reporting
- Create charts/graphs for trends
- Generate summary statistics
- Present findings in structured format

Key Metrics to Calculate

Commit Metrics

Total commits
Commits per day/week/month
Average commits per active period
Longest streak of daily commits
Periods of inactivity

Contributor Metrics

Total unique contributors
Top N contributors (typically top 5-10)
Contribution percentage per contributor
One-time vs. recurring contributors
New contributors over time

Release Metrics

Total releases
Time between releases (min, max, average)
Release frequency trend
Semantic versioning patterns
Pre-release vs. stable releases

Output Formats

Summary Report

Repository: owner/repo
Analysis Period: YYYY-MM-DD to YYYY-MM-DD

Commit Activity:
- Total Commits: N
- Active Contributors: N
- Average Commits/Week: N
- Most Active Period: YYYY-MM

Top Contributors:
1. Name (N commits, X%)
2. Name (N commits, X%)
...

Recent Releases:
- v1.2.3 (YYYY-MM-DD) - N days since previous
- v1.2.2 (YYYY-MM-DD) - N days since previous
...

Detailed Analytics

Time-series data in CSV/JSON format
Visualization-ready datasets
Contributor breakdown by time period
Release calendar with annotations

Common Patterns & Tips

Handle rate limiting: Always check API rate limit headers and implement exponential backoff

Large repositories: For repos with 10k+ commits, consider:

Analyzing recent history only (e.g., last 12 months)
Sampling commits rather than processing all
Using shallow clones for git-based analysis

Privacy considerations: GitHub API exposes public data only; private repos require authentication

Timezone handling: Normalize all timestamps to UTC for consistent analysis

Bot commits: Filter out automated commits (dependabot, renovate) for human contributor analysis

Email normalization: Same contributor may use different email addresses; consider consolidation

Error Handling

Repository not found: Verify owner/repo spelling
Rate limit exceeded: Implement retry logic or use authentication
Empty history: Check if repository has been initialized
API changes: GitHub API versioning may affect endpoints

Agent Skills: GitHub Repository Analysis

Install this agent skill to your local

Skill Files