Agent Skills Hub is a security-reviewed directory of 1,213+ MCP servers and AI agent skills, each graded A–F before installation — purpose-built for teams that need to evaluate tool safety and workflow fit without relying on package names alone.
What is Agent Skills Hub?
Agent Skills Hub is a curated registry that turns public repository signals, permission context, install guidance, and editorial review into searchable pages. It accepts MCP server and Claude Code skill metadata from GitHub and outputs security-graded listings with installation instructions and workflow-fit analysis. The platform runs as a web directory and is maintained by an independent team (no company name stated on the page).
Key Features
- Security grading (A–F) — Every skill is scanned for 45+ attack vectors including arbitrary code execution, API key leakage, and unbounded file access before it appears in search results.
- 1213+ total skills — Covers official MCP servers, OpenClaw skills, and community-contributed tools across categories like development, e-commerce, finance, and marketing.
- OpenClaw Skills Leaderboard — Ranked by "heat score" (community adoption + security audit score) from 2,993 cleaned skills; includes install counts and star ratings.
- Trending picks and scenario pages — Weekly editor-curated top skills, plus dedicated guides for browser automation, MCP discovery, and agent workflow libraries.
- Value-added guides — Separate landing pages for "Agent Skills Hub", "Cursor Skills Hub", "MCP Agent Skills", and "Skills Library for Agents" to give search engines targeted citation targets instead of routing all queries through the homepage.
- Vulnerability intelligence — Public threat intelligence table showing real vulnerability patterns detected in the registry (e.g., RCE, credential leaks, path traversal).
Who is it for?
- AI agent developers building with Claude Code, Cursor, or Codex who need to shortlist MCP servers before adding them to a local or production workflow.
- Security-conscious engineering teams that want pre-vetted skills with visible risk grades (A–F) and explicit permission scope documentation.
- Technical decision-makers evaluating skill libraries for internal use — the site offers scenario blueprints (e.g., e-commerce automation, smart expense audit) and category-based collections.
What can you do with Agent Skills Hub?
- Search and compare skills by security grade — Use the directory filters to find only Grade A skills (979 verified) or drill into specific categories like "Developer Toolkit" (580 skills) or "All Verified (Grade A)".
- Explore trending skills weekly — The site publishes a "Weekly Trend Pulse" covering MCP registries, workflow automation, browser agents, memory systems, and research agents.
- Deploy solution blueprints — Featured workflows (e.g., "E-commerce Automation", "Smart Expense Audit") include install paths and security posture notes, ready for one-click deployment.
How does it work?
- Browse the skill directory via search or category filters.
- Review each skill's security grade, permission scope, source evidence, and install path on its detail page.
- Use the provided installation guidance (often MCP JSON config or CLI commands) to add the skill to your agent environment.
- For premium verification, submit a skill for manual static analysis ($99 audit) to receive a "VERIFIED SECURE" badge and priority routing in the directory.
Pricing
Agent Skills Hub is free to browse and use — all listed skills are accessible without payment. A premium security audit and verification badge costs $99 per skill and includes manual static analysis plus a security certificate.
FAQ
