Skills Janitor

Skills Janitor: Audit, track usage, and manage your Claude Code skills efficiently.

Introduction:

Skills Janitor is a powerful plugin for Claude Code designed to audit, track usage, and compare your AI skills ecosystem. It features 9 focused skills with zero external dependencies, allowing developers to maintain a clean and organized workspace. Users can identify duplicate skills, check for errors, and generate comprehensive health reports. With built-in GitHub search and market comparison tools, Skills Janitor ensures you are using the best available tools while identifying unused or broken skills to optimize your developer workflow.

Added On:

2026-04-15

Monthly Visitors:

634285.6K

Code & IT

Skills Janitor - AI Tool Screenshot and Interface Preview

Skills Janitor Product Information

Skills Janitor: The Essential Plugin for Claude Code Skill Management

Maintaining a clean and efficient development environment is crucial for productivity. Skills Janitor is a specialized plugin designed specifically for Claude Code that allows users to audit, track usage, and compare their skills ecosystem. With a focus on organization and health, Skills Janitor provides 9 dedicated actions to ensure your Claude Code skills remain high-performing and clutter-free.

Built with zero dependencies, Skills Janitor operates using Bash and Python 3, making it a lightweight yet powerful addition to your developer toolkit. Whether you are looking to find duplicate skills, fix errors, or discover new tools on GitHub, Skills Janitor offers a comprehensive suite of commands to manage your AI-driven workflow.

What's Skills Janitor?

Skills Janitor is an audit and management plugin for Claude Code. It serves as a "janitor" for your digital workspace, keeping your skills organized by providing deep insights into how they are used. The tool helps developers identify which skills are active, which are broken, and which are redundant.

By parsing conversation history, Skills Janitor can generate usage reports that highlight your most-used skills and recommend the removal of unused ones. It also facilitates skill discovery by allowing you to search and compare your current local skills against those available in the wider GitHub ecosystem.

Features

Skills Janitor is packed with features designed to optimize your Claude Code experience:

Comprehensive Auditing: View a full inventory of all installed skills.
Duplicate Detection: Easily identify redundant skills that perform identical tasks.
Health Checks & Auto-Fixing: Scan for errors or missing information and automatically resolve issues with a safe preview mode.
Usage Tracking: Analyze your conversation history to see which skills you actually invoke and which have been dormant.
Skill Discovery: Search GitHub for new skills to install and compare them against your current set.
Zero Dependencies: Requires no pip installs or node_modules, relying strictly on Bash, Python 3, and curl.
Safety First: Destructive operations are dry-run by default and never delete content without explicit confirmation.

How to Use Skills Janitor

Skills Janitor integrates seamlessly into your workflow using slash commands with dedicated autocomplete or natural language triggers.

Installation

To install Skills Janitor, you can use the recommended plugin marketplace command:

/plugin marketplace add khendzel/skills-janitor
/plugin install skills-janitor

Alternatively, you can clone the repository directly into your Claude directory:

git clone https://github.com/khendzel/skills-janitor ~/.claude/skills/skills-janitor

Command Reference

Once installed, you can use the following commands to manage your Claude Code skills:

/janitor-audit: Displays a full inventory of all your skills.
/janitor-usage: Shows a detailed report of which skills you invoke versus which remain unused.
/janitor-search: Searches GitHub for new Claude Code skills based on keywords.
/janitor-compare: Performs market analysis by comparing your skill against alternatives on GitHub.
/janitor-check: Scans your skills for errors and missing information.
/janitor-fix: Automatically fixes detected skill problems (previews changes first).
/janitor-report: Generates a full health report of your entire skill ecosystem.
/janitor-cleanup: Identifies and removes broken skills.
/janitor-duplicates: Finds duplicate skills that perform overlapping functions.

Usage Tracking Examples

Skills Janitor provides actionable insights. For instance, a usage report might look like this:

=== Skills Janitor - Usage Report === Period: 4 weeks Active skills: 4 / 36 (11%) Unused skills: 32 (89%) Recommendation: Remove 32 unused skills

Use Case

Developer Workflow Optimization

Developers using Claude Code often accumulate many skills over time. Skills Janitor is ideal for those who want to prune their environment to ensure only the most efficient and error-free tools are active.

Skill Maintenance

If you are a skill creator, you can use /janitor-check and /janitor-fix to ensure your metadata is correct, and /janitor-compare to see how your skill stacks up against others in the GitHub community.

Resource Discovery

Users looking to expand their capabilities can use /janitor-search to find high-quality skills on GitHub. By setting a GITHUB_TOKEN, users can enjoy higher rate limits and more accurate search results.

FAQ

Q: Does Skills Janitor delete my files automatically? A: No. Skills Janitor never deletes anything without explicit confirmation and uses dry-runs by default for all potentially destructive operations.

Q: Can it modify marketplace plugins? A: No, the tool is designed to respect your ecosystem and will not modify plugin or marketplace-managed skills.

Q: What are the system requirements? A: You only need Bash, Python 3, and curl. There is no need for external package managers like pip or npm.

Q: How does the usage tracking work? A: It parses your local conversation history to identify which skill commands have been invoked over a specific period.

Alternatives Tools

Edgee Codex Compressor

Edgee Token Compression: Optimize Codex Costs and Context Efficiency for Coding Agents

Edgee is a cutting-edge AI gateway and compression layer designed to optimize Codex and LLM performance. By reducing redundant context and fresh token intake, Edgee helps developers slash API costs by over 35% and cut input token usage by nearly half. Through its advanced compression-lab benchmarks, Edgee demonstrates how to maintain high-quality model output while significantly improving cache hit rates and workload efficiency. Ideal for engineering teams using agentic coding tools, Edgee eliminates context bloat without changing developer workflows.

Code & IT

LaReview

LaReview: The AI-Powered Code Review Workbench for Senior Engineers and High-Signal Feedback

LaReview is a local-first, open-source code review workbench designed to transform messy code changes into structured review plans. Unlike standard AI bots that generate comment spam, LaReview focuses on a 'Plan First' philosophy, helping engineers understand system impacts through flow-based reviews rather than file-by-file analysis. It integrates seamlessly with existing AI coding agents like Claude, Gemini, and Mistral, offering features such as AI-powered planning, visual diagrams, and local context via linked Git repos. With built-in CLI support and Git host sync for GitHub and GitLab, LaReview ensures zero data leaks by processing information locally. It empowers teams to prioritize 'merge confidence' by identifying authentic bugs, enforcing custom architectural rules, and learning from rejected feedback to eliminate nitpicks. LaReview is free, open-source under MIT/Apache 2.0, and compatible with macOS, Linux, and WSL.

Code & IT

AgentPulse by Rectify

Rectify: The AI-Powered Agent Platform for Monitoring, Support, Roadmaps, and Automated Debugging

Rectify is a comprehensive operations platform powering over 1,000 products, designed for founders and run by AI agents. By centralizing monitoring, analytics, support, roadmaps, and changelogs, Rectify allows teams to control their entire workspace through conversation. Featuring Quanta, an AI Operations Agent, Rectify automates complex tasks like code scanning, session replays, and support drafting. It bridges the gap between user frustration and developer solutions by transforming vague bug reports into actionable data, complete with annotated screenshots and console logs. With built-in uptime monitoring, visual feedback tools, and a powerful browser extension, Rectify provides a unified debugging experience that integrates seamlessly with GitHub and AI coding assistants.

Code & IT

Mercury Edit 2

Mercury Edit 2: The Fastest Diffusion-Based LLM for Next-Edit Prediction and Real-Time Coding

Mercury Edit 2 is a purpose-built diffusion Large Language Model (dLLM) designed for the most latency-sensitive coding workflows. By generating tokens in parallel, Mercury Edit 2 provides lightning-fast next-edit predictions that feel instantaneous. This advanced model upgrades previous iterations to deliver higher quality, more targeted code suggestions with a 48% higher acceptance rate. Optimized for the Inception Platform and integrated with editors like Zed, Mercury Edit 2 offers superior speed and quality across multiple programming languages and development scenarios.

Code & IT

Google Gemma 4

Gemma 4: Google's Most Intelligent Open Models for Advanced Reasoning and Agentic Workflows

Gemma 4 represents a breakthrough in open AI technology, offering four versatile model sizes including Effective 2B, Effective 4B, 26B MoE, and 31B Dense. Built on Gemini 3 research, Gemma 4 delivers industry-leading intelligence-per-parameter, high-performance reasoning, and native multimodal support for vision, audio, and video. Released under the Apache 2.0 license, it empowers developers to build autonomous agents and mobile-first applications with complete data sovereignty and seamless hardware integration.

Code & IT

Cursor 3

Cursor 3: The Unified AI Workspace for Building Software with Autonomous Agents

Cursor 3 is a revolutionary software development platform designed for the third era of AI coding. Built from scratch, it moves beyond manual file editing to a unified agent-first interface. With Cursor 3, developers can manage multiple agents in parallel across different repositories, seamlessly hand off tasks between local and cloud environments, and utilize Composer 2—a frontier coding model. The platform features an integrated browser, a dedicated marketplace for plugins and MCPs, and a streamlined UX for managing the entire lifecycle from commit to merged PR. Whether you are using the Cursor IDE or the new autonomous agent workspace, Cursor 3 provides the foundational tools—including LSPs and full file visibility—to ensure high-velocity, high-quality software development in an AI-driven world.

Code & IT

Cosyra

Cosyra: A Powerful Cloud Terminal for AI Coding and Development on Mobile Devices

Cosyra is a revolutionary mobile development environment that brings a full Linux terminal to your phone. It comes pre-installed with powerful AI coding agents like Claude Code, Codex CLI, OpenCode, and Gemini CLI, allowing developers to ship code from anywhere. With 30 GB of persistent storage and support for Node.js, Python, and Git, Cosyra provides an isolated Ubuntu container for seamless mobile coding. Pro features include session hibernation and 120 hours of monthly access, making it the ultimate tool for reviewing PRs, refactoring code, and managing repositories on the go.

Code & IT

GLM-5V-Turbo

GLM-5V-Turbo: Z.AI's Advanced Multimodal Coding Foundation Model

GLM-5V-Turbo is Z.AI’s premier multimodal coding foundation model designed for vision-based coding and agentic workflows. Supporting image, video, and text inputs, it features a 200K context length and 128K output tokens. It excels at complex coding tasks, GUI exploration, and autonomous action execution, integrating seamlessly with agents like Claude Code and OpenClaw through native multimodal fusion and reinforcement learning.

Code & IT

Loading related products...