Installable components for AI coding assistants -- skills, MCP servers, agent loops, subagents, hooks, plugins, prompts, and CLI tools, each with a source link and an install command.
600 installable components across 8 categories. Every entry is sourced from a real open-source project. Works with Claude Code, OpenCode, Codex, Cursor, Gemini CLI, Copilot, and 30+ AI coding assistants.
94 skills across 9 categories: Development (31) · Productivity (17) · Design (11) · Content (10) · DevOps (8) · Marketing (6) · Data (5) · Testing (4) · Security (2)
Full catalog: skills/ · catalog.json
| Name | Description | Source |
|---|---|---|
| Executing Plans | Executes implementation plans with verification at each step. | obra/superpowers |
| Frontend Design | Creates distinctive, production-grade frontend interfaces with high design quality. | anthropics/skills |
| Grill Me | Get relentlessly interviewed about a plan or design until every branch of the decision tree is resolved. | mattpocock/skills |
| Vercel React Best Practices | React and Next.js performance optimization guidelines from Vercel Engineering. 40+ rules across 8 categories. | vercel-labs/agent-skills |
| Find Skills | Discover and install new skills for your agent. | vercel-labs/skills |
133 mcps across 19 categories: Developer Tools (15) · AI & Machine Learning (14) · Agent Orchestration (14) · Communication (11) · Databases (10) · Search (10) · DevOps (8) · Security (7) · Official Reference (6) · Research & Data (6) · Browser Automation (5) · Cloud Platforms (5) · Marketing (5) · Monitoring (5) · Finance (4) · Design (3) · Blockchain (3) · Data Engineering (1) · Mobile (1)
Full catalog: mcps/ · catalog.json
| Name | Description | Source |
|---|---|---|
| n8n MCP | Fair-code workflow automation platform with native AI capabilities and 400+ integrations | n8n-io/n8n |
| MarkItDown | Converts PDF, Office, HTML and image files into clean Markdown for AI consumption | microsoft/markitdown |
| Browser Use MCP | AI-powered browser automation with 100K+ stars. Enables agents to browse the web, fill forms, extract data, and interact with web pages autonomously. | browser-use/browser-use |
| Filesystem MCP | Secure file read/write/list/search operations with configurable directory access controls | modelcontextprotocol/servers |
| Netdata MCP | Real-time infrastructure monitoring with MCP integration for system metrics | netdata/netdata |
103 loops across 7 categories: engineering (53) · multi-agent (14) · evaluation (11) · meta (10) · operations (6) · design (6) · content (3)
Full catalog: loops/ · catalog.json
| Name | Description | Source |
|---|---|---|
| The docs sweep | Keeps documentation aligned with the current codebase | Forward-Future/loopy |
| Daily Triage Pattern | 1d-2h maintenance loop for triaging issues, PRs, and priorities | cobusgreyling/loop-engineering |
| Issue Burn-Down Line | Curated backlog through worker, QA, and checkpoint stations | rohanbalkondekar/agent-loop-patterns |
| Alpha Loop | Agent-agnostic automated development loop: Plan -> Build -> Test -> Review -> Ship | bradtaylorsf/alpha-loop |
| The Workflow Loop | Complete shipping loop for coding agents with Plan/Execute/Land/Learn phases | iKon85/agent-workflow-kit |
31 subagents across 5 categories: Subagent Collection (15) · Agent Harness (6) · Official SDK (4) · Platform Format (4) · Curated Directory (2)
Full catalog: subagents/ · catalog.json
| Name | Description | Source |
|---|---|---|
| oh-my-openagent | Agent harness with async subagent orchestration. Manages subagent delegation, parallel execution, and result aggregation across multiple agent personas. | code-yeongyu/oh-my-openagent |
| Microsoft Agent Framework | Microsoft's comprehensive framework for building autonomous AI agents with multi-agent patterns, Azure integration, and .NET/Python support. | learn.microsoft.com |
| awesome-claude-code | Ecosystem directory indexing 47K+ stars of Claude Code resources including subagent collections, skills, MCP servers, and community extensions. | hesreallyhim/awesome-claude-code |
| awesome-claude-code-subagents | Curated list of 154+ Claude Code subagent definitions including code-reviewer, debugger, test-writer, planner, and more. VoltAgent's definitive collection of reusable agent personas. | VoltAgent/awesome-claude-code-subagents |
| oh-my-codex | Codex orchestration harness with 33 agent prompts. Manages subagent lifecycle, handoffs, and coordinated task execution for Codex CLI. | Yeachan-Heo/oh-my-codex |
23 hooks across 7 categories: Security (4) · Automation (4) · Session Management (4) · Quality (3) · Safety (3) · Utility (3) · Notifications (2)
Full catalog: hooks/ · catalog.json
| Name | Description | Source |
|---|---|---|
| Secret Scanner | Scans for leaked secrets (API keys, tokens, passwords) before writing or editing files | rohitg00/awesome-claude-code-toolkit |
| Block Dangerous Commands | Blocks rm -rf, fork bombs, curl|sh, and other destructive shell commands | karanb192/claude-code-hooks |
| Lasso Security Prompt Injection Defenses | Enterprise-grade prompt injection detection and prevention hooks for Claude Code. Detects and blocks indirect prompt injection attacks via tool outputs. | lasso-security/claude-hooks |
| awesome-claude-code-hooks | Curated directory of Claude Code hooks organized by category. Meta-resource for discovering hooks in the ecosystem. | ithiria894/awesome-claude-code-hooks |
51 plugins across 9 categories: Claude Code (11) · OpenCode (9) · Cross-Tool (8) · VS Code AI (6) · Cursor (5) · Windsurf (4) · JetBrains (4) · Copilot (3) · Aider (1)
Full catalog: plugins/ · catalog.json
| Name | Description | Source |
|---|---|---|
| Superpowers Plugin | Comprehensive skill pack with 60+ pre-built skills for Claude Code workflows | superpowers.dev |
| Official Claude Code Marketplace | Official marketplace for Claude Code plugins with 255+ community-contributed extensions | claude.ai |
| Awesome CursorRules | Curated collection of cursor rules by PatrickJS with 20K stars | PatrickJS/awesome-cursorrules |
| AgentEnv | Project-scoped AI agent and plugin environment manager. Manages agent configurations across multiple AI coding tools. | kvcache-ai/AgentENV |
| Flutter Agent Plugins | Official Flutter team agent plugins bundling Flutter skills, rules, and Dart/Flutter MCP config for widget tests, layout, routing, and architecture. | docs.flutter.dev |
84 prompts across 17 categories: Coding (9) · Architecture (8) · Image Generation (8) · DevOps (6) · General Purpose (6) · Code Review (5) · Debugging (5) · Documentation (5) · Reasoning (5) · Testing (4) · Data (4) · Design (4) · Productivity (4) · Agent Workflows (4) · Security (3) · Content (3) · Marketplace (1)
Full catalog: prompts/ · catalog.json
| Name | Description | Source |
|---|---|---|
| Senior Code Reviewer | Thorough code review with security, performance, and maintainability analysis | f/prompts.chat |
| Prompt Engineer | Meta-prompt for designing and optimizing other prompts | dair-ai/Prompt-Engineering-Guide |
| Craft Code Refactor | Systematic code improvement with test coverage maintenance | hoangatg/awesome-ai-prompts-2026 |
| SQL Query Optimizer | Analyze and optimize SQL queries with index suggestions | openai/openai-cookbook |
| Image Generation Prompt Crafter | Craft detailed prompts for DALL-E, Midjourney, and Flux | Ryan-yang125/gpt-image-2-prompts |
81 tools across 14 categories: AI Coding CLIs (14) · Code Analysis (9) · Cloud & DevOps (7) · Formatting & Linting (7) · Agent Training & Eval (7) · Git Utilities (6) · Package Managers (5) · Docker & Containers (5) · API Testing (5) · Database CLIs (4) · Monitoring (4) · Agent Memory (3) · Terminal Enhancement (3) · AI APIs (2)
Full catalog: tools/ · catalog.json
| Name | Description | Source |
|---|---|---|
| Ollama | Run LLMs locally. Llama, Mistral, Phi, Gemma, and more. | ollama/ollama |
| Gemini CLI | Google's AI coding agent in the terminal. | google-gemini/gemini-cli |
| Claude Code | Anthropic's agentic coding tool. Terminal-native AI pair programmer. | anthropics/claude-code |
| Bun | Incredibly fast JavaScript runtime, bundler, and package manager. | oven-sh/bun |
| Netdata | Real-time performance and health monitoring. | netdata/netdata |
We welcome contributions! You can:
- Manual PR -- fork, add an entry to a
catalog.jsonfile, validate, and submit. See contributing.md. - Agent-automated -- give your AI agent the contributing.md guide and it handles everything.
- Open an issue -- suggest a tool we should add.
All data comes from catalog.json files in each folder. These catalogs are the single source of truth for programmatic discovery.