🔌

MCP Tools

Open-source tools and servers for the Model Context Protocol (MCP) ecosystem. Connect AI agents to databases, browsers, file systems, and APIs through a unified protocol — includes official and community-built MCP servers.

186 projects

System Prompts and Models of AI Tools

143.3k · Unknown
Active A+

Full system prompts, internal tools and AI models from 40+ popular AI tools including Cursor, Devin, Windsurf, Manus, Lovable, and more.

prompt-engineeringagenttools +2
  • · System prompt collection — Complete system prompts from 40+ AI tools (Cursor, Devin, Windsurf, Manus, etc.)
  • · AI model intelligence — Records internal AI model names and versions used by each tool for tech stack insights
  • · Internal tool definitions — Exposes registered function/tool schemas revealing agent capabilities

Awesome MCP Servers

93.9k · Unknown
Active A+

A curated collection of hundreds of community-verified MCP server implementations spanning databases, search engines, dev tools, browser automation, and more, helping developers quickly discover and integrate MCP services for their use cases.

mcptoolsapi +1
  • · Most comprehensive MCP server directory - hundreds of community-verified server implementations, continuously updated
  • · Multi-language coverage - includes TypeScript, Python, Go, Rust, C#, Java, Ruby and more
  • · Well-organized categories - 30+ categories including databases, browser automation, dev tools, search, file systems

MCP Servers

90.0k · TypeScript
Active A+

MCP Servers provides a large collection of reusable Model Context Protocol server implementations, giving agents standardized tool capabilities.

mcptoolsintegration +1
  • · Official reference implementations - authoritative server collection maintained by the MCP steering group showcasing protocol features and SDK usage
  • · Multi-language SDK support - 10 SDKs available: TypeScript, Python, Go, Rust, Java, Kotlin, C#, Ruby, Swift, PHP
  • · Core reference servers - Everything (testing), Fetch (web scraping), Filesystem, Git, Memory (knowledge graph), Sequential Thinking, Time

Awesome Claude Skills

74.3k · Python
Active A

A curated list of awesome Claude Skills, resources and tools for customizing Claude AI workflows.

mcptoolspython +1
  • · 1000+ production-ready skills - Covering 9 categories including document processing, dev tools, data analysis, marketing, and writing
  • · Cross-agent compatible - Works with Claude Code, Claude.ai, OpenAI Codex, Cursor, Gemini CLI, Windsurf and more
  • · Progressive loading architecture - Each skill loads only name and description (~100 tokens), full content on demand, hundreds coexist

Headroom

68.6k · Python
Active A+

Context compression layer that compresses tool outputs, logs, files, and RAG chunks before they reach the LLM, saving 60-95% of tokens while preserving answer quality. Available as library, proxy, or MCP server.

token-compressionoptimizationmcp +2
  • · Compress everything before LLM receives it, reducing 60-95% of tokens
  • · Three usage modes: library, proxy, and one-command agent wrapping
  • · Built-in MCP server with cross-agent shared memory

Last30Days Skill

61.0k · Python
Active A+

An AI agent skill that researches any topic across Reddit, X, YouTube, HN, Polymarket and the web, then synthesizes a grounded summary.

mcptoolspython +2
  • · Multi-platform parallel search — Search Reddit, X, YouTube, TikTok, HN, Polymarket and 12+ platforms simultaneously
  • · Social signal scoring — Ranked by real user engagement (likes, shares, bet amounts) not SEO
  • · AI agent synthesis — Extracts key information from multi-platform results into structured summary reports

CLIProxyAPI

50.0k · Go
Active A+

Wrap Gemini CLI, Antigravity, ChatGPT Codex, Claude Code as an OpenAI/Gemini/Claude/Codex compatible API service for unified agent access.

goapiagent +2
  • · OpenAI-compatible API — Wraps Claude Code, Codex, Gemini CLI as standard OpenAI-format REST endpoints
  • · OAuth login — Use Claude Code, Codex, Grok Build subscriptions directly via OAuth flow
  • · Multi-account round-robin — Load balancing across multiple accounts for Gemini, OpenAI, Claude, Grok

Obsidian Skills

47.8k · Unknown
Normal A

Agent skills for Obsidian. Teach your agent to use Markdown, Bases, JSON Canvas, and use the CLI.

agentcodingtools +2
  • · Obsidian Markdown skill — Teach agents to create and edit Obsidian Flavored Markdown with wikilinks, embeds, callouts, properties
  • · Obsidian Bases skill — Create and edit .base files with views, filters, formulas and summaries
  • · JSON Canvas skill — Create and edit JSON Canvas files with nodes, edges, groups and connections

Marketing Skills

46.6k · JavaScript
Active A+

Marketing skills for Claude Code and AI agents, covering CRO, copywriting, SEO, analytics and growth engineering.

mcptoolsjavascript +1
  • · SEO Skill Suite — Complete SEO workflow covering audits, AI SEO, schema markup, site architecture, and content strategy
  • · CRO & Conversion — Covers signup flows, popups, paywalls, A/B testing, and other conversion rate optimization scenarios
  • · Copywriting & Content — Full-spectrum content skills for copy, editing, cold email, social media, video, and image copy

Antigravity Awesome Skills

45.9k · Python
Active A+

An installable library of 1,400+ agentic skills for Claude Code, Cursor, Codex CLI, Gemini CLI and more, with installer CLI, bundles and workflows.

mcptoolspython +2
  • · 1500+ installable skills — Covers dev, test, security, infrastructure, product, and marketing domains
  • · npm installer — One-command install via npx antigravity-awesome-skills to your tool's skills directory
  • · Multi-tool compatible — Supports Claude Code, Cursor, Codex CLI, Gemini CLI, Kiro, OpenCode, Copilot

Codebase Memory MCP

41.9k · C
Active A+

High-performance code intelligence MCP server that indexes codebases into a persistent knowledge graph, supporting 66 languages with sub-millisecond queries and 99% fewer tokens.

mcpknowledge-graphcode-analysis +2
  • · Extreme indexing speed: Linux kernel (28M LOC) indexed in 3 minutes via RAM-first pipeline
  • · 159 languages supported through vendored tree-sitter grammars in a single binary
  • · Hybrid LSP semantic type resolution for Python, TypeScript, PHP, C#, Go, C, and C++

Claude Code Plugins

35.8k · Python
Active A

The official Anthropic-managed directory of high-quality Claude Code plugins, installable via the /plugin command.

claude-codepluginsmarketplace +2
  • · Official plugin marketplace - directory of Claude Code plugins curated and reviewed by Anthropic
  • · Dual sources - bundles both internal Anthropic plugins and vetted third-party partner plugins
  • · One-line install - load any plugin with the /plugin install command inside Claude Code

GitHub MCP Server

32.7k · Go
Active A+

GitHub's official MCP Server providing standardized access to GitHub APIs for AI agents, supporting repository management, issue handling, and PR operations.

mcpgithubgo
  • · Repository Management — Browse code, search files, analyze commits, understand project structure across any repo
  • · Issue & PR Automation — Create, update, manage issues and PRs. AI triages bugs and reviews code changes
  • · CI/CD Intelligence — Monitor GitHub Actions workflows, analyze build failures, manage releases, get pipeline insights

Vibe-Trading

32.3k · Python
Active A+

Vibe-Trading is an open-source personal trading agent stack that combines LLMs, MCP tooling, and multi-agent workflows for algorithmic trading and backtesting.

ai-agentalgorithmic-tradingbacktesting +5
  • · One command empowers your trading agent with full quantitative trading capabilities via LLM + MCP toolchain
  • · Backtesting engine supporting 7 data sources (Tushare, OKX, CCXT, AkShare, etc.) with local caching
  • · Connector-first architecture supporting IBKR, Robinhood, Tiger, Alpaca, OKX, Binance paper trading

Vercel Agent Skills

30.8k · JavaScript
Active A

Vercel's official collection of agent skills, providing practical skill modules and tools for AI coding agents.

agentcodingtools +2
  • · Vercel project optimization — Auto-audit project cost, performance, caching, function usage, and billing
  • · React best practices — 40+ performance optimization rules from Vercel engineering
  • · Web design audit — 100+ rules checking accessibility, performance, and UX

Google Workspace CLI

30.7k · Rust
Active A

A unified CLI for Google Workspace covering Drive, Gmail, Calendar, Sheets, Docs and more, with built-in AI agent skills for automation.

cliagentrust +2
  • · Dynamic command discovery — Reads Google Discovery Service at runtime to auto-build all API commands; new Google endpoints auto-supported
  • · 40+ agent skills — Built-in AI agent skills with structured JSON output for LLMs to manage Workspace directly
  • · Multiple auth methods — Supports interactive OAuth, Service Account, pre-obtained tokens, CI/headless credential export

Serena

28.8k · Python
Active A+

Powerful MCP toolkit for coding that provides semantic retrieval and editing capabilities, serving as an IDE for AI agents

coding-agentmcplanguage-server +3
  • · Semantic code retrieval — symbol-level code understanding with relational structure, no need to read entire files
  • · Cross-file refactoring — symbol-level rename, move, and reference lookup completed in a single atomic call
  • · 40+ language support — LSP backend supports Python, Java, TypeScript, Go, Rust and 40+ other languages

FastMCP

27.5k · Python
Active A+

FastMCP is a fast, Pythonic library for building MCP servers and clients with over 1 million daily downloads, making it easy to create Model Context Protocol tools.

mcpserverprotocol +2
  • · Decorator-Based Tools — Create MCP tools with Python functions and @mcp.tool decorator, auto-generating schema and docs
  • · MCP Client — Connect to any MCP server (local or remote) with full protocol handshake and session management
  • · Interactive Apps — Generate interactive UIs for tools that render directly in conversations

Blender MCP

26.7k · Python
Active A+

MCP server for Blender that enables AI agents to directly control the 3D modeling software for natural language-driven scene creation, model manipulation, and rendering automation.

mcptoolsagent +1
  • · Bidirectional communication — Blender addon creates socket server, MCP Server bridges AI and Blender
  • · 3D object manipulation — create, modify, delete 3D objects in Blender via natural language
  • · Material and rendering control — natural language application of materials, colors, lighting, render settings

MCP Python SDK

24.2k · Python
Active A+

MCP Python SDK is the official Python implementation for building MCP servers and agent-side integrations with a standardized tool protocol.

mcppythonsdk +1
  • · FastMCP high-level interface — quickly define MCP Server tools, resources and prompts with decorators
  • · Multi-transport support — supports stdio, SSE and Streamable HTTP standard transport methods
  • · Structured output — tool return values automatically serialized into structured JSON responses

n8n MCP

22.8k · TypeScript
Active A+

An MCP for Claude Desktop, Claude Code, Windsurf, and Cursor to build n8n workflows through natural language interaction.

mcpworkflowautomation +2
  • · 1851 n8n node coverage — Complete docs and properties for 822 core + 1029 community nodes
  • · MCP protocol server — Provides n8n knowledge to Claude Desktop, Cursor, Windsurf and other AI IDEs
  • · Template search — 2352 workflow templates filterable by complexity, role, and time

Context Mode

20.3k · TypeScript
Active A

Context Mode is a context window optimization tool for AI coding agents that sandboxes tool output for 98% context reduction across 12 major platforms.

context-optimizationmcpclaude-code +2
  • · Sandboxes tool output, reducing 315KB raw data to 5.4KB — a 98% context reduction
  • · SQLite + FTS5 session continuity that recovers key context via BM25 search after conversation compaction
  • · Think in Code paradigm: LLM writes analysis scripts instead of processing raw data, saving 100x context

Agentic

18.1k · TypeScript
Stale B

Turn any API into a paid MCP service instantly. Helps developers wrap existing APIs into MCP-compatible agent tools for AI agent capability extension.

mcpapiagent +2
  • · Curated MCP marketplace — hand-vetted tools with comprehensive integration tests and evals
  • · First-class MCP support for both publishing and consuming agent tools
  • · HTTP to MCP compatibility layer bridging existing APIs to the MCP protocol

(24 / 186)

Related Articles

rate limitingcost controlprompt injection

Agent Rate Limiting and Cost Attack Defense: Token Quotas, Sliding Windows, Attack Vector Practice

Systematic guide to rate limiting and cost attack defense for LLM Agents in production: single-user token quotas, multi-tenant sliding windows, prompt injection amplifying cost, DDoS protection, and integration with LangChain/CrewAI/AutoGen frameworks.

AI Agent安全Prompt Injection

AI Agent Security in Practice: From Prompt Injection to Defense in Depth

A systematic walkthrough of three major attack surfaces in AI agents, with practical code examples for prompt injection defense, tool permission scoping, and output filtering.

MCPModel Context ProtocolClaude Code

Best MCP Servers in 2026: 25 Servers Worth Integrating, From Official to Community

A 2026 review of the 25 Model Context Protocol (MCP) servers most worth integrating, covering GitHub, databases, browsers, filesystem, code search, Google Workspace, Slack, Notion, and more. Includes official SDK vs community implementations and configuration in Claude Code, Cursor, and Cline.

browser-useWeb AutomationAgent

Web Automation Agent in Practice: Limits and Best Practices of browser-use

A practical breakdown of browser-use strengths and limits in web task automation, with strategies for stable execution and failure recovery.

上下文工程长上下文RAG

Context Engineering: Context Decay and Recovery in Long-Conversation Agents

Long-conversation agents fail at context management, not model capability. A systematic comparison of sliding window, retrieval injection, and layered compression strategies with practical decay diagnosis and recovery patterns.

deep-researchagent-architectureiterative-retrieval

Deep Research Agents in Practice: From Single-Shot Search to Iterative Reasoning

Decomposes five open-source deep-research projects across three sub-stages with reusable query-rewriting code.