Agent Leaderboard

The go-to leaderboard for discovering trending AI agent ecosystem repositories on GitHub — covering Agent Skills, MCP Servers, Prompt Libraries, AI Frameworks, and Auto Research tools, all ranked by GitHub stars and updated automatically.

中文文档

---

Five Boards, One Place

Board	Repos Tracked	What It Covers
🛠 Agent Skills	500+	Installable skills/plugins for Claude Code, Cursor, Copilot, Codex and other AI coding agents
🔌 MCP Servers	1,500+	Model Context Protocol servers that extend agent tool-use capabilities
📚 Prompt Library	1,100+	Curated system prompts, prompt collections, and prompt engineering guides
🏗 AI Frameworks	1,200+	Agent orchestration frameworks, multi-agent platforms, and LLM app scaffolding
🔬 Auto Research	100+	Autonomous deep-research agents and AI-powered research tools

→ Explore all boards at agentskills.media

---

Top 10 Agent Skills

#	Repository	Description	Stars	Category
1	affaan-m/everything-claude-code	The agent harness performance optimization system. Skills, instincts, memory,…	★ 185,355	claude
2	NousResearch/hermes-agent	The agent that grows with you	★ 154,273	claude
3	x1xhlol/system-prompts-and-models-of-ai-tools	FULL Augment Code, Claude Code, Cluely, CodeBuddy, Comet, Cursor, Devin AI…	★ 137,532	claude
4	multica-ai/andrej-karpathy-skills	A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy	★ 133,529	claude
5	mattpocock/skills	Skills for Real Engineers. Straight from my .claude directory.	★ 88,185	claude
6	nextlevelbuilder/ui-ux-pro-max-skill	An AI SKILL that provides design intelligence for building professional UI/UX	★ 79,571	claude
7	bytedance/deer-flow	An open-source long-horizon SuperAgent harness that researches, codes, and creates	★ 68,076	other
8	JuliusBrussee/caveman	🪨 why use many token when few token do trick — Claude Code skill that cuts 65% of tokens	★ 61,217	claude
9	ComposioHQ/awesome-claude-skills	A curated list of awesome Claude Skills, resources, and tools	★ 60,260	claude
10	safishamsi/graphify	AI coding assistant skill (Claude Code, Codex, OpenCode, Cursor, Gemini CLI…)	★ 48,702	claude

→ See full Agent Skills ranking at agentskills.media

Top 10 MCP Servers

#	Repository	Description	Stars	Category
1	punkpeye/awesome-mcp-servers	A collection of MCP servers	★ 87,171	general
2	sansan0/TrendRadar	AI-driven public opinion & trend monitor with multi-platform data	★ 57,880	dev_tools
3	ChromeDevTools/chrome-devtools-mcp	Chrome DevTools for coding agents	★ 40,041	web
4	microsoft/playwright-mcp	Playwright MCP server	★ 32,726	web
5	github/github-mcp-server	GitHub's official MCP Server	★ 29,976	official
6	modelcontextprotocol/servers	Model Context Protocol Servers (official reference implementations)	★ —	official
7	jlowin/fastmcp	The fast, Pythonic way to build MCP servers	★ —	dev_tools
8	Cursor-to-API/cursor-api	Use Cursor as an OpenAI-compatible API	★ —	dev_tools
9	bgauryy/octocode-mcp	MCP server for semantic code research and context generation	★ —	dev_tools
10	PatrickJS/awesome-cursorrules	Configuration files that enhance Cursor AI editor experience	★ —	general

→ See full MCP Servers ranking at agentskills.media/#mcp

Top 10 Prompt Libraries

#	Repository	Description	Stars	Category
1	f/awesome-chatgpt-prompts	Awesome ChatGPT Prompts — share, discover and collect prompts	★ 162,535	collection
2	x1xhlol/system-prompts-and-models-of-ai-tools	Extracted system prompts from major AI tools	★ 137,850	system
3	dair-ai/Prompt-Engineering-Guide	Guides, papers, and resources for prompt engineering	★ 74,758	engineering
4	PlexPt/awesome-chatgpt-prompts-zh	ChatGPT 中文调教指南	★ 60,191	collection
5	obra/superpowers	An agentic skills framework & software development methodology	★ 198,224	engineering
6	anthropics/anthropic-cookbook	Recipes and notebooks for using Claude effectively	★ —	engineering
7	openai/openai-cookbook	Examples and best practices for working with the OpenAI API	★ —	engineering
8	brexhq/prompt-engineering	Tips and tricks for working with Large Language Models	★ —	engineering
9	linexjlin/GPTs	Leaked prompts of GPTs	★ —	system
10	ai-boost/awesome-prompts	Curated list of awesome ChatGPT prompts	★ —	collection

→ See full Prompt Library ranking at agentskills.media/#prompts

Top 10 AI Frameworks

#	Repository	Description	Stars	Category
1	obra/superpowers	An agentic skills framework & software development methodology	★ 198,224	general
2	langchain-ai/langchain	The agent engineering platform	★ 137,134	orchestration
3	vllm-project/vllm	High-throughput and memory-efficient inference and serving for LLMs	★ 80,489	memory
4	TauricResearch/TradingAgents	Multi-Agents LLM Financial Trading Framework	★ 77,314	multi_agent
5	FoundationAgents/MetaGPT	The Multi-Agent Framework: First AI Software Company	★ 68,127	multi_agent
6	microsoft/autogen	A programming framework for agentic AI	★ —	orchestration
7	crewAIInc/crewAI	Framework for orchestrating role-playing, autonomous AI agents	★ —	orchestration
8	pydantic/pydantic-ai	Agent Framework / shim to use Pydantic with LLMs	★ —	orchestration
9	BerriAI/litellm	Call all LLM APIs using the OpenAI format	★ —	routing
10	openai/swarm	Educational framework for lightweight multi-agent orchestration	★ —	orchestration

→ See full AI Frameworks ranking at agentskills.media/#frameworks

Top 10 Auto Research Tools

#	Repository	Description	Stars	Category
1	karpathy/autoresearch	AI agents running research on single-GPU nanochat training automatically	★ 81,483	general
2	bytedance/deer-flow	An open-source long-horizon SuperAgent harness that researches, codes, and creates	★ 68,076	deep_research
3	microsoft/qlib	AI-oriented Quant investment platform	★ 43,098	literature
4	666ghj/BettaFish	多Agent舆情分析助手，打破信息茧房，还原舆情原貌	★ 40,936	deep_research
5	khoj-ai/khoj	Your AI second brain. Self-hostable. Get answers from the web or your docs.	★ 34,578	deep_research
6	stanford-oval/storm	An LLM-powered knowledge curation system that researches and generates wiki-like articles	★ 28,228	deep_research
7	assafelovic/gpt-researcher	An autonomous agent that conducts deep research on any data using any LLM	★ 27,113	deep_research
8	mvanhorn/last30days-skill	AI agent skill that researches any topic across Reddit, X, YouTube, HN, Polymarket	★ 26,015	deep_research
9	virattt/dexter	An autonomous agent for deep financial research	★ 25,789	data_research
10	dzhng/deep-research	An AI-powered research assistant that performs iterative, deep research on any topic	★ 18,924	deep_research

→ See full Auto Research ranking at agentskills.media/#research

---

Features

Feature	Description
🌐 Five Boards	Agent Skills · MCP Servers · Prompt Library · AI Frameworks · Auto Research
🔍 Multi-filter	Keyword search, language, Stars threshold (All / 500+ / 1k+ / 5k+ / 10k+), time range
🏷️ Use-case chips	Click tags to filter by use case — multi-select with OR logic
⊞ Dual views	Grid / List view toggle, 24 / 48 / 96 items per page
❤️ Favorites	Bookmark repos; persisted in localStorage
🌐 i18n	Switch between English and Chinese with one click
🌙 Theme	Dark mode (default) / Light mode toggle
📊 Auto tagging	Parses repo description + topics to assign use-case labels
📄 Pagination	Smart pagination with ellipsis

Quick Start

# 1. Clone the repo
git clone https://github.com/jaychempan/Agent-Leaderboard.git
cd Agent-Leaderboard

# 2. Fetch data (Python 3.8+, no extra dependencies)
python3 scripts/fetch_data.py           # Agent Skills
python3 scripts/fetch_mcp.py            # MCP Servers
python3 scripts/fetch_prompts.py        # Prompt Library
python3 scripts/fetch_frameworks.py    # AI Frameworks
python3 scripts/fetch_auto_research.py  # Auto Research

# Optional: supply a GitHub Token to raise the API rate limit (60 → 5000 req/hr)
python3 scripts/fetch_data.py --token ghp_xxxxxxxxxxxx

# 3. Serve locally
# Option A — with live reload (auto-refreshes browser on file change, requires Node)
npx live-server --port=8080

# Option B — basic static server (no live reload, no extra dependencies)
python3 -m http.server 8080
# then open http://localhost:8080

MCP Skill Discovery

Use the catalog directly from any AI chat client that supports MCP. This Bash command installs the server, creates an isolated venv under ~/.local/share/skills-discovery-mcp, and configures detected clients:

curl -fsSL https://raw.githubusercontent.com/jaychempan/Agent-Leaderboard/main/scripts/install.sh \
  | SKILLS_DISCOVERY_CONFIGURE_CLIENTS=auto bash

Use macOS, Linux, WSL, or Git Bash for the installer. Native PowerShell does not support this Bash syntax.

For Codex, verify the server is registered, then start a new Codex session:

codex mcp list

Quick examples are in docs/mcp-quick-use.zh.md.

To skip client config or target specific clients:

SKILLS_DISCOVERY_CONFIGURE_CLIENTS=none bash scripts/install.sh
SKILLS_DISCOVERY_CONFIGURE_CLIENTS=codex,claude,cursor bash scripts/install.sh

The installer backs up config files before editing them. If your client is not configured automatically, add this fallback manually:

{
  "mcpServers": {
    "skills-discovery": {
      "command": "skills-discovery-mcp"
    }
  }
}

Re-run scripts/install.sh to update an existing install. To uninstall:

curl -fsSL https://raw.githubusercontent.com/jaychempan/Agent-Leaderboard/main/scripts/uninstall.sh | bash

You can also run the server directly from a cloned checkout:

python3 -m mcp.skills_discovery.server

By default the server reads the daily remote catalog at https://agentskills.media/data/discovery_index.json, generated from the leaderboard data. To test a local or custom index, point it at an absolute file URL:

SKILLS_DISCOVERY_INDEX_URL=file:///absolute/path/to/data/discovery_index.json \
python3 -m mcp.skills_discovery.server

Example chat queries:

• Find Codex testing skills
• Recommend Claude UI/UX skills
• Top MCP browser automation servers

The server returns guidance, matching repositories, and links only. It does not execute install commands or modify your environment.

File Structure

agent-leaderboard/
├── index.html                       # Single-page app (all five boards)
├── favicon.svg                      # Site logo / favicon
│
├── src/
│   ├── shared.css                   # Shared styles (dark / light theme)
│   └── shared.js                    # Shared logic (i18n, filtering, rendering, routing)
│
├── scripts/
│   ├── fetch_data.py                # Crawl Skills repos → data/data.js
│   ├── fetch_mcp.py                 # Crawl MCP repos → data/mcp_data.js
│   ├── fetch_prompts.py             # Crawl Prompt repos → data/prompts_data.js
│   ├── fetch_frameworks.py         # Crawl Framework repos → data/frameworks_data.js
│   ├── fetch_auto_research.py       # Crawl Research repos → data/auto_research_data.js
│   ├── build_discovery_index.py     # Build MCP discovery index → data/discovery_index.js
│   └── fetch_utils.py               # Shared fetch utilities
│
└── data/
    ├── data.json / data.js                       # Agent Skills data
    ├── mcp_data.json / mcp_data.js               # MCP Servers data
    ├── prompts_data.json / prompts_data.js        # Prompt Library data
    ├── frameworks_data.json / frameworks_data.js  # AI Frameworks data
    ├── auto_research_data.json / auto_research_data.js  # Auto Research data
    └── discovery_index.json / discovery_index.js  # Daily MCP discovery index

Automated Updates (GitHub Actions)

Each board has its own daily workflow that keeps data fresh automatically.

Workflow	Schedule	Board
update-skills.yml	Daily	Agent Skills
update-mcp.yml	Daily	MCP Servers
update-prompts.yml	Daily	Prompt Library
update-frameworks.yml	Daily	AI Frameworks
update-research.yml	Daily	Auto Research
weekly-update.yml	On demand	All boards (manual trigger)

Setup (one-time):

1. Go to Settings → Secrets and variables → Actions in your GitHub repo.
2. Add a secret named GH_TOKEN with a Personal Access Token (public repo read access is sufficient).
3. Push this repo to GitHub. Workflows run on schedule automatically.

You can also trigger a full refresh manually via Actions → Update All (Manual) → Run workflow.

Inclusion Criteria

	Skills	MCP Servers	Prompt Library	AI Frameworks	Auto Research
Min stars	★ 100+	★ 50+	★ 50+	★ 50+	★ 100+
Keywords	claude / codex / cursor / copilot / skill	mcp / model context protocol / tool server	prompt / system prompt / chatgpt prompts	langchain / autogen / crewai / framework	deep research / autonomous / literature / RAG
Categories	Claude · Codex · Cursor · Copilot · Other	Official · Web · Dev Tools · API · General	Collection · System · Engineering · General	Orchestration · Multi-Agent · Memory · Routing	Deep Research · Web Research · Literature · Data · Knowledge Base

Tech Stack

• Pure HTML + CSS + JS — no framework, no build step
• Python 3 stdlib only — urllib only, no pip install needed
• GitHub REST Search API — public endpoint, token optional

Contributing

PRs welcome to improve search queries, use-case tag rules, or the UI!