code-analysis

Analyze Git repositories and generate developer evaluation reports in Markdown, HTML, JSON, and PDF.

39|1|Updated Jul 2, 2026
One-click install
npx skills add https://github.com/HKU-MMLab/UniClawBench --skill code-analysis-hku-mmlab
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: code-analysis
Source: https://github.com/HKU-MMLab/UniClawBench/tree/main/injection/101_skill_usage/task_101_27_pr_triage/skills/code-analysis-skills
Command: npx skills add https://github.com/HKU-MMLab/UniClawBench --skill code-analysis-hku-mmlab

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires gitpython, radon, pylint, pydriller, tabulate, jinja2, pyyaml, click, reportlab, and includes scripts (resource) and references (resource) components.

What problem does it solve? Understanding developer behavior and code quality from raw Git history requires manual digging through logs. This Skill scans Git repositories and produces structured, multi-dimensional reports covering commit patterns, work habits, efficiency, code style, code quality, slacking index, and graded developer evaluations. ## Core Features & Use Cases - Multi-Dimensional Git Analysis: Measures commit frequency, message quality, weekend/late-night coding ratios, code churn, rework rate, Bus Factor, Conventional Commits compliance, bug-fix ratio, and Python cyclomatic complexity via radon. - Developer Evaluation & Leaderboards: Computes a weighted six-dimension score (S/A/B/C/D/E/F grades) with strengths, weaknesses, actionable suggestions, plus a 0-100 slacking index and team rankings. - Flexible Scanning & Output: Analyze a single repo or recursively scan a directory of repos, filter by author, date range, or branch, and export reports as Markdown, JSON, HTML, or PDF. - Use Case: A team lead wants to compare Alice and Bob's code quality over the past quarter; the Skill scans the repo, filters by author and date range, and outputs a side-by-side HTML report with scores and improvement suggestions. ## Quick Start Ask the agent to analyze the Git repository at a given path and generate a developer evaluation report, optionally specifying authors, a date range, and an output format such as HTML or PDF.

Frequently Asked Questions about code-analysis

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I analyze a Git repository's developer activity in Python?▼

Run the CLI with python -m src.main -r /path/to/repo to analyze all contributors. Use -a to filter specific authors, -s and -u for date ranges, and -f to choose markdown, json, html, or pdf output.

How to compare code quality between multiple developers?▼

Pass multiple author flags such as -a "Alice" -a "Bob" when running the analysis. The report generates side-by-side comparison tables, a developer score leaderboard, and a slacking index leaderboard across all dimensions.

Can I scan multiple Git repositories under one directory at once?▼

Yes, add the --scan-all flag to recursively discover all .git repositories under the given path. Scanning defaults to a maximum depth of 5 levels and skips bare repositories and symlink loops.

Does the code complexity analysis work for languages other than Python?▼

No, cyclomatic complexity analysis uses the radon library and only applies to .py files. Other quality metrics like bug-fix ratio, revert frequency, and large-commit ratio work for any language tracked in Git history.

Why is PDF report generation failing or missing?▼

PDF generation prefers weasyprint, then falls back to pdfkit, and finally to reportlab. If none of these are available in the environment, PDF output fails; use markdown, json, or html formats instead.

What are the limitations of Git-based developer evaluation?▼

Evaluations rely solely on commit history and miss non-coding work like reviews, meetings, and design. Roles such as architects naturally commit less, so scores and the slacking index should be interpreted in context, not used for punitive decisions.