What is Agent Skills Hub
Agent Skills Hub is a security-graded directory for Claude skills and MCP servers. It indexes over 120,000 open-source AI agent tools, including MCP servers, Claude skills, Codex skills, AI coding assistants, and agent tools. Every entry is security-graded, quality-scored, and refreshed every 8 hours. The platform helps developers find safe, production-ready tools without guessing.
How to use Agent Skills Hub
- Browse by Category: Explore tools organized by primary function, such as MCP Server, Claude Skill, Codex Skill, Agent Tool, Prompt Library, AI Coding Assistant, or AI Tool.
- Search by Keyword: Use the search bar to find specific tools or skills.
- Filter by Programming Language: Narrow down results based on language preferences.
- Explore Curated Scenarios: Visit scenario pages like MCP Browser Automation, Code Review, or Web Scraping to see top-ranked tools for specific use cases.
- Compare Tools: Use the Compare feature to evaluate multiple tools side by side.
- Use the Skill Analyzer: Check security grades and platform compatibility for informed decisions.
Features of Agent Skills Hub
- Security Grading: Every entry is security-graded to help users identify safe tools.
- Quality Scoring: Each tool receives a composite quality score from 0 to 100 based on six dimensions: completeness, clarity, specificity, examples, README structure, and agent readiness.
- Auto-Updated Every 8 Hours: Data refreshes automatically to show the latest tools.
- 10 Weighted Signals: Scoring includes GitHub stars, code quality, documentation, community activity, and more.
- Category and Scenario Browsing: Tools are organized by category and curated scenario pages.
- Compare Feature: Side-by-side comparison of multiple tools.
- Skill Analyzer: Provides security grades and platform compatibility checks.
- Open Source: The entire source code is open under MIT on GitHub.
- Transparent Methodology: Every scoring dimension is documented, and data snapshots are archived.
Use Cases of Agent Skills Hub
- Developers building with AI agents can find and compare MCP servers, Claude skills, and automation tools.
- Teams evaluating open-source tools can use quality scores and security grades to make informed decisions.
- Researchers tracking the AI agent ecosystem can access a comprehensive, up-to-date directory.
- Developers looking for production-ready tools can filter by quality score and security grade.
- Anyone needing browser automation, code review, web scraping, or workflow automation can explore curated scenario pages.
FAQ
Q: How often is the data updated? A: Data refreshes every 8 hours.
Q: How are tools scored? A: Each tool receives a composite quality score from 0 to 100 based on six dimensions: completeness, clarity, specificity, examples, README structure, and agent readiness. The score also incorporates 10 weighted signals including GitHub stars, code quality, documentation, and community activity.
Q: Is the platform open source? A: Yes, the entire source code is open under MIT on GitHub.
Q: How are tools discovered? A: Tools are automatically discovered from GitHub using multi-phase data collection: search, enrichment, README analysis, and quality scoring.
Q: Can I compare tools? A: Yes, the Compare feature allows side-by-side evaluation of multiple tools.
Q: Is there a newsletter? A: Yes, a weekly newsletter delivers trending AI agent tools, new MCP servers, and top skill picks every Monday.
