Code Analysis Skill

📦 GitHub: https://github.com/Wscats/code-analysis-skills

Scan specified repositories or directories containing Git repositories, analyze and compare developers' commit habits, work patterns, development efficiency, code style, code quality, and slacking index. Provide blunt, data-driven evaluations for each developer with scores, grades, strengths, weaknesses, and actionable suggestions. Output structured reports in Markdown / HTML / JSON / PDF.

💬 Natural Language (Recommended)

You don't need to memorize any commands — simply describe what you need in your own language:

🇺🇸 English

CODEBLOCK0

🇨🇳 中文

CODEBLOCK1

🇯🇵 日本語

CODEBLOCK2

🇰🇷 한국어

CODEBLOCK3

🇪🇸 Español

CODEBLOCK4

🇫🇷 Français

CODEBLOCK5

🇩🇪 Deutsch

CODEBLOCK6

Note: You need to provide the repository path (repo_path) — it is a required parameter. If you are already working within a repository context, the agent may infer the path from the conversation, but an explicit path is always recommended for accuracy.

The Skill understands all the languages above. Just describe what you need and it will run the analysis on your repository and return a structured report.

---

🚀 Quick Start (CLI)

Install Dependencies

CODEBLOCK7

For higher quality PDF output (optional):

CODEBLOCK8

Common Commands

CODEBLOCK9

CLI Parameters

Parameter	Short	Description	Default
INLINECODE1	INLINECODE2	Path to Git repository or parent directory	Required
INLINECODE3

| Recursively scan all .git repositories | false | | --author | -a | Filter by author (repeatable) | All authors | | --since | -s | Start date (ISO format) | None | | --until | -u | End date (ISO format) | None | | --branch | -b | Branch to analyze | Active branch | | --format | -f | Output format: markdown, json, html, pdf (comma-separated for multiple) | markdown | | --output | -o | Output file path | stdout |

---

Use Cases

• - Analyze developer behavior in a Git repository
• Compare team members' commit habits and development efficiency
• Understand code quality trends and style consistency
• Batch-analyze all .git repositories under a directory
• Generate work habit reports (active hours, weekend/late-night coding, streaks, etc.)
• Evaluate each developer's overall capability with scores, strengths, weaknesses, and suggestions
• View the team's "Slacking Index" leaderboard
• Produce formal PDF reports for review

Workflow

Step 1: Confirm Analysis Parameters

Ask the user for the following information:

• - Repository path: A single Git repo path, or a parent directory containing multiple repos
• Scan scope: Whether to scan all .git repos under the directory (--scan-all)
• Target authors: Analyze specific developers (multi-select) or all contributors
• Date range: Optional start/end dates (ISO format, e.g., 2024-01-01)
• Branch: Branch to analyze; defaults to the current active branch
• Output format: markdown (default), json, html, pdf, or comma-separated combination

Step 2: Run the Analysis

Execute the analysis script with the confirmed parameters (see Quick Start above for command examples).

Step 3: Interpret the Report

The report covers seven dimensions. Walk the user through the key findings for each:

1. 1. 🏆 Developer Evaluation — Overall score (S/A/B/C/D/E/F), strengths, weaknesses, improvement suggestions, one-line verdict
2. 🐟 Slacking Index — Activity level, trivial commit ratio, disappearance ratio, low output, procrastination signals
3. 📝 Commit Habits — Commit frequency, commit size, merge ratio, message quality
4. ⏰ Work Habits — Active hour distribution, weekend/late-night coding ratio, consecutive coding streaks
5. 🚀 Development Efficiency — Code churn rate, rework rate, Bus Factor, file ownership
6. 🎨 Code Style — Language distribution, Conventional Commits compliance, file classification
7. 🔍 Code Quality — Bug fix ratio, revert frequency, large commit ratio, test coverage, complexity

For multi-developer analysis, additional sections include:

• - 📋 Cross-comparison summary table
• 🏆 Developer score leaderboard
• 🐟 Slacking Index leaderboard

Step 4: Deep-Dive into Evaluation Results

For each developer's evaluation, deliver a blunt, no-nonsense interpretation to the user:

1. 1. Score & Grade: Total score (0-100) and corresponding grade (S/A/B/C/D/E/F)
2. Six Dimension Scores: Commit discipline, work consistency, efficiency, code quality, code style, engagement
3. Strengths: Each backed by concrete data, not generic praise
4. Weaknesses: No sugarcoating — point directly at the problem and its impact
5. Suggestions: Actionable improvement measures, each immediately executable
6. Slacking Index: Interpret each signal and its meaning

Available Resources

Scripts

• - src/main.py — Main entry point with CLI argument support, orchestrates the full analysis pipeline and generates reports
• INLINECODE32 — Repository scanner, discovers single or recursively scans multiple Git repositories
• INLINECODE33 — Base analyzer class providing Git history traversal and author filtering
• INLINECODE34 — Commit habit analysis (frequency, size, message quality)
• INLINECODE35 — Work habit analysis (active hours, weekends, late nights, streaks)
• INLINECODE36 — Development efficiency analysis (churn, rework, bus factor)
• INLINECODE37 — Code style analysis (language distribution, commit conventions)
• INLINECODE38 — Code quality analysis (bug fixes, reverts, complexity)
• INLINECODE39 — Slacking index analysis (activity, trivial commits, disappearance patterns, procrastination, etc.)
• INLINECODE40 — Developer evaluation engine (overall scoring, strengths/weaknesses, suggestions, verdicts)
• INLINECODE41 — Markdown report generator
• INLINECODE42 — JSON report generator
• INLINECODE43 — HTML report generator (with rich visual styling)
• INLINECODE44 — PDF report generator (supports weasyprint/pdfkit/reportlab fallback)

Reference Documents

• - references/metrics-guide.md — Metric definitions, calculation methods, and healthy value reference ranges. Read this file when users ask about the meaning of a specific metric.

⚠️ Privacy & Data Security Notice

Important: This tool extracts personal developer activity data from Git commit history, including but not limited to:

• - Commit timestamps (down to the hour)
• Weekend/late-night coding frequency
• Individual commit frequency and output volume
• Code ownership attribution
• Slacking index and behavioral assessments

Before using, you must adhere to the following principles:

1. 1. Informed Consent — Obtain informed consent from all relevant developers before analyzing their repositories
2. Non-Punitive Use — Analysis results must not be directly used for performance reviews, compensation decisions, or punitive management
3. Contextual Understanding — Data must be interpreted within actual work context (e.g., architects naturally commit less; that does not indicate low output)
4. Data Protection — Generated reports contain personal information and should be securely stored, not publicly shared
5. Compliance — Ensure usage complies with your organization's HR policies and local data protection regulations (e.g., GDPR)
6. Local Execution — This tool runs entirely locally and does not transmit any data to external servers

Evaluation System

Overall Score (0-100)

Grade	Score Range	Meaning
S	90-100	Top-tier contributor, excellent across all dimensions
A

80-89 | Outstanding developer, reliable and efficient | | B | 70-79 | Solid contributor with minor room for improvement | | C | 60-69 | Adequate, but needs improvement in multiple areas | | D | 50-59 | Barely passing, has clear weaknesses | | E | 35-49 | Below expectations, requires serious attention | | F | 0-34 | Critical issues, needs coaching or intervention |

Six Dimension Weights

Dimension	Weight	What It Evaluates
📝 Commit Discipline	15%	Commit frequency, message quality, convention compliance
⏰ Work Consistency

15% | Routine regularity, work continuity | | 🚀 Efficiency | 20% | Code churn rate, rework rate, output volume | | 🔍 Code Quality | 25% | Bug fix rate, revert rate, test coverage, complexity | | 🎨 Code Style | 10% | Conventional Commits, issue references | | 💪 Engagement | 15% | Inverse of slacking index signals |

Slacking Index (0-100)

Level	Score Range	Meaning
🔥 Workaholic	0-20	Highly engaged, continuous contributions
✅ Normal

21-40 | Healthy work pattern | | 😏 Suspicious | 41-60 | Some slacking signals detected | | 🐟 Slacking Pro | 61-80 | Significant low-engagement indicators | | 🏆 Slacking Master | 81-100 | Professional-grade slacking |

Notes

• - Analyzing large repositories (100K+ commits) may take a long time; consider limiting the date range
• Python code complexity analysis depends on the radon library and only works on .py files
• Author matching supports fuzzy matching (matches on name or email containing the keyword)
• Directory scanning defaults to a maximum depth of 5 levels to avoid excessive recursion
• PDF generation prefers weasyprint, falls back to pdfkit, and ultimately falls back to reportlab
• Evaluation results are based solely on Git commit history and do not represent a developer's full capability
• The slacking index is for reference only and should be interpreted in actual work context
• This tool runs entirely locally and does not send data to any external server
• Always obtain informed consent before analyzing team repositories
• Report results must not be directly used for performance reviews or punitive management decisions

... EOF