research

Devscope Research Dataset

This directory contains authority analysis data from running Devscope on notable open-source repositories. The dataset serves as a benchmark for code health metrics and validates Devscope's analysis capabilities across diverse codebases.

📊 Dataset Overview

Each analyzed repository includes:

• summary_compact.txt - Single-line health summary (grade, metrics, scan time)
• report.md - Full markdown analysis report with detailed metrics
• metadata.json - Structured data including:
  • Repository name and URL
  • Analysis timestamp
  • Scan duration
  • Key metrics (grade, files, lines, test ratio, languages)
  • Health scores and risk assessment

🎯 Methodology

Repository Selection Criteria

Repositories are selected based on:

1. Popularity - High star count and active community
2. Diversity - Multiple programming languages and domains
3. Maturity - Established projects with stable codebases
4. Open Source - Publicly accessible for reproducibility

Analysis Configuration

All analyses use:

CodebaseAnalyzer(
    repo_path,
    detect_git=True,        # Enable git metrics
    enable_intelligence=True # Enable advanced analysis
)

Features enabled:

• Complexity analysis (cyclomatic, cognitive)
• Test detection and ratio calculation
• Git metrics (commit frequency, age)
• Hotspot identification
• Maintainability scoring
• Risk assessment
• Onboarding difficulty evaluation

Metrics Collected

Metric	Description	Range
Maintainability Grade	Overall code health score	A (best) to F (worst)
Risk Level	Code maintenance risk	Low / Medium / High
Onboarding Difficulty	New developer learning curve	Easy / Moderate / Hard
Test Ratio	Test-to-code line ratio	0.0 - 3.0+
Complexity	Average cyclomatic complexity	1.0+
Scan Time	Analysis duration	Seconds

📁 Dataset Structure

research/
├── README.md                    # This file
├── analyze_repo.py              # Script for analyzing GitHub repos
├── analyze_self.py              # Script for analyzing devscope itself
│
├── devscope_self_analysis/      # Devscope analyzing itself
│   ├── metadata.json
│   ├── report.md
│   └── summary_compact.txt
│
├── facebook_react/              # React library analysis
│   ├── metadata.json
│   ├── report.md
│   └── summary_compact.txt
│
└── index.json                   # Summary index of all analyses

🚀 Running Your Own Analysis

Analyze a GitHub Repository

# Using the provided script
python research/analyze_repo.py

# Or manually analyze a specific repo
devscope scan /path/to/repo --markdown > output.md

Analyze Devscope Itself

python research/analyze_self.py

Analyze Custom Repository

from pathlib import Path
from datetime import datetime
from devscope.analyzer import CodebaseAnalyzer
from devscope.formatters import generate_markdown_summary
import json

# Analyze repository
start = datetime.now()
analyzer = CodebaseAnalyzer(
    Path("/path/to/repo"),
    detect_git=True,
    enable_intelligence=True
)
result = analyzer.analyze()
scan_time = (datetime.now() - start).total_seconds()

# Generate report
report = generate_markdown_summary(result)
print(report)

# Save metadata
metadata = {
    "repo_name": "my-repo",
    "analyzed_at": datetime.now().isoformat(),
    "scan_time_seconds": scan_time,
    "maintainability_grade": result.health_score.maintainability_grade,
    # ... additional fields
}

📈 Research Applications

This dataset can be used for:

1. Benchmarking - Compare your project against industry standards
2. Validation - Verify Devscope's analysis accuracy
3. Research - Study code health patterns across languages
4. Documentation - Real-world examples of analysis output
5. Training - Learn what good/bad metrics look like

🔬 Example Analysis: Devscope Self-Analysis

Repository: EhsanAzish80/Devscope

Quick Stats:

• Grade: A
• Files: 51 Python files
• Lines: 10,399 LOC
• Test Ratio: 0.84 (excellent test coverage)
• Risk: Low
• Scan Time: 0.19s

Key Findings:

• Well-tested codebase with strong coverage
• Modern Python practices (type hints, dataclasses)
• Low complexity averages
• Comprehensive test suite
• Active development (recent commits)

See full analysis: devscope_self_analysis/report.md

📝 Data Format Specification

metadata.json Schema

{
  "repo_name": "owner/repository",
  "repo_url": "https://github.com/owner/repository",
  "analyzed_at": "2026-02-13T12:00:00.000000",
  "scan_time_seconds": 1.23,
  "metrics": {
    "total_files": 100,
    "total_lines": 50000,
    "maintainability_grade": "B",
    "overall_score": 85.5,
    "risk_level": "Low",
    "test_ratio": 0.75
  },
  "languages": {
    "Python": 45.2,
    "JavaScript": 30.1,
    "HTML": 24.7
  }
}

🤝 Contributing New Analyses

To add your repository analysis to this dataset:

1. Run analysis using analyze_repo.py or manual script
2. Verify output includes all required files
3. Document findings in your repository folder
4. Update index.json with new entry
5. Submit PR with descriptive commit message

Requirements for inclusion:

• Public repository (for reproducibility)
• Notable project (1000+ stars or significant community impact)
• Clean analysis (no errors or warnings)
• Complete metadata

📊 Dataset Statistics

Last Updated: February 13, 2026

Repository	Grade	Files	Lines	Language	Scan Time
devscope	A	51	10,399	Python	0.19s
(Add more as analyzed)	-	-	-	-	-

🔗 Related Resources

• Devscope Documentation
• Analysis Methodology
• Scoring System
• GitHub Action

📄 License

This research dataset is provided under the same license as the Devscope project. Individual repositories analyzed remain under their original licenses.

🙏 Acknowledgments

Thanks to all open-source projects included in this dataset. Your work makes the software community better.

---

Note: Analysis results are point-in-time snapshots. Repository health may change as code evolves. Re-run analyses periodically for up-to-date metrics.