Skills

Agent skills for the OpenFinance backend that teach AI agents how to trade on Polymarket and Hyperliquid, bridge assets via Relay, manage on-chain wallets, and handle fiat on-ramps. Skills reference actual backend routes and WebSocket channels for direct HTTP/WS, SDK, or MCP integration.

A collection of skills for AI agents to integrate with the Openfort blockchain platform, providing embedded wallet setup across React/React Native/Swift/Unity, backend wallet operations on EVM and Solana, fee sponsorship policies, and CLI/MCP tools.

PeerBanHelper is an open-source personal network firewall that automatically bans unwanted, leeching, and abnormal BitTorrent clients by connecting to supported apps' Web APIs and identifying potentially threatening connections.

A plugin marketplace and discovery platform for Claude Code that bundles 117 agents, 175 commands, 28 hooks, 26 skills, and 51 plugins, along with community plugin indexing and a web-based browsing interface.

A plugin marketplace and discovery platform for Claude Code offering curated agents, commands, hooks, skills, and plugins, plus community plugin discovery with a browsable web UI at buildwithclaude.com.

Part of Build with Claude, a plugin marketplace and discovery platform for Claude Code offering 117 agents, 175 commands, 28 hooks, and 51 bundled plugin packages, along with community discovery indexing over 20,000 plugins and 4,500 MCP servers.

A skill from the Build with Claude plugin marketplace, a discovery platform for Claude Code that includes 117 specialized agents, 175 slash commands, 28 hooks, 26 skills, and 51 plugin packages covering the full development lifecycle.

A plugin marketplace and discovery platform for Claude Code that bundles 117 specialized agents, 175 slash commands, 28 hooks, and 51 plugins across categories like DevOps, security, and language-specific development. Includes a web UI for browsing, searching, and one-click installation of all plugin types.

A Membrane app integration skill that teaches AI agents to connect with external applications, handling OAuth, API keys, and credential management automatically via the Membrane CLI.

from cline/skills

from refoundai/lenny-skills

A comprehensive AI-powered development toolkit that extends Claude Code with specialized agents for dozens of frameworks (React, Angular, Vue, Python, Rust, Go, and more), MCP servers, a web dashboard with visual orchestration, code generator, project templates, and an Electron desktop app.

Part of amplihack, a development framework that adds structured workflows, persistent memory, specialized agents, and quality gates to Claude Code and GitHub Copilot CLI. Features a 22-step systematic workflow, knowledge graph persistence across sessions, and self-improvement through reflection and pattern capture.

from garethmanning/education-agent-skills

Pattern critic agent skill that reviews pattern code for violations of documented rules, gotchas, and anti-patterns, producing categorized checklists with PASS/FAIL results.

An integration skill from the Membrane platform that enables AI agents to connect with third-party apps, providing app-specific actions, authentication management, and drop-in support for Claude Code, Cursor, GitHub Copilot, and other agents.

Payment integration skill from the GoodVibes plugin for implementing payment processing workflows with specialized agents.

A skill for programmatic PDF processing that extracts text and tables, creates PDFs, merges and splits documents, and fills forms using pypdf and command-line tools.

A Chinese-localized version of Anthropic's official Claude skills repository, covering document processing (Word, PDF, PPTX, Excel), creative design, frontend development, and MCP server building, with all documentation fully translated to Simplified Chinese.

Part of NVIDIA's official verified agent skills catalog, providing portable instruction sets that teach AI agents to correctly use NVIDIA CUDA-X libraries, AI Blueprints, and platform tools.

Official NVIDIA-verified agent skills that teach AI agents to use NVIDIA CUDA-X libraries, AI Blueprints, and platform tools correctly. Skills are mirrored from NVIDIA product repos and cover areas like numerical optimization (cuOpt), performance tuning, and MoE dispatcher selection.

Official NVIDIA-verified skills that teach AI agents how to use NVIDIA CUDA-X libraries, AI Blueprints, and platform tools correctly. Skills are maintained in product repos and mirrored daily via an automated sync pipeline.

The largest open-source cybersecurity skills library for AI agents, featuring 754 skills across 26 security domains. Compatible with 26+ platforms and 5 frameworks for comprehensive security assessment and education.

Part of the largest open-source cybersecurity skills library for AI agents, featuring 754 production-grade skills across 26 security domains with 5 framework mappings. Compatible with 26+ AI platforms including Claude Code.

Part of Anthropic Cybersecurity Skills, the largest open-source cybersecurity skills library for AI agents, with 24 Red Teaming skills covering full-scope engagements, Active Directory attacks, and phishing simulation.

The largest open-source cybersecurity skills library for AI agents, featuring 754 skills across 26 security domains with compatibility for 5 frameworks and 26+ platforms.

Audits PHP source code for encryption and key security issues, detecting weak hashing (MD5/SHA1), insecure encryption modes, JWT/HMAC signature verification flaws, and hardcoded keys, with severity-rated findings, PoC, and remediation guidance.

A PHP source code audit skill for CSRF vulnerabilities that analyzes state-changing interfaces for CSRF protection, traces token generation and validation flows, and outputs exploitability ratings with PoC examples and remediation advice.

A PHP source code audit skill for detecting arbitrary file read and path traversal vulnerabilities. Identifies file read sinks, traces path sources, analyzes bypass techniques, and outputs exploitability ratings with PoC and remediation guidance.

A PHP source code audit skill that identifies NoSQL injection vulnerabilities by tracing user input into MongoDB and DocumentDB query construction, detecting operator injection ($gt, $ne, $where), and providing severity ratings with PoC and remediation advice.