awesome-awesome
MCP server that helps discover and rank libraries from awesome lists by tech stack and category, using a state-machine workflow to find relevant sections and present ranked repos.
README
awesome-awesome
English | 中文
Discover and rank libraries from awesome lists — MCP server + Skill.
awesome-awesome helps you pick libraries for a specific technology stack.
Give it a tech and a category (e.g. "vue" + "icon"), and it will:
- Find the matching awesome-list(s)
- Locate the relevant section
- Collect every GitHub repo under that section
- Rank them by stars → last-update → forks → watchers (archived projects excluded)
- Present a table for you to choose from
It's a state-machine: you (or your AI client) make repeated discover calls,
picking an awesome-list, then a section, until you get a ranked candidate list.
Install
Prerequisites
- Node.js 20+ (or Bun 1.2+)
- A GitHub Personal Access Token with
public_reposcope — generate one here
For AI clients (MCP)
Add to your MCP client config (e.g., Claude Code's mcp.json):
{
"mcpServers": {
"awesome-awesome": {
"command": "npx",
"args": ["-y", "github:awesome-awesome/awesome-awesome/packages/mcp"],
"env": { "GITHUB_TOKEN": "ghp_your_token_here" }
}
}
}
Then install the Skill so your AI knows how to drive the tools:
# Ask your AI client to call the install_skill tool, or run manually:
GITHUB_TOKEN=ghp_xxx npx -y github:awesome-awesome/awesome-awesome/packages/mcp --install-skill
For humans (direct)
git clone https://github.com/awesome-awesome/awesome-awesome
cd awesome-awesome
bun install
Usage
Once configured, just ask your AI client naturally:
"Find me a Vue icon library"
The Skill will drive the discover tool, ask you to pick from any ambiguous
choices, and return a ranked table.
Tools
| Tool | Purpose |
|---|---|
discover |
Main state-machine entry — finds candidates or asks for choices |
list_sections |
Inspect an awesome-list's sections manually |
get_repo_details |
Deep stats on one repo (for final comparison) |
install_skill |
Copy the Skill files to your local skills directory |
Development
bun install
bun test # unit tests
GITHUB_TOKEN=ghp_xxx bun test:integration
bun run lint
See docs/superpowers/specs/2026-07-21-awesome-awesome-design.md for the
full design and docs/superpowers/plans/2026-07-21-mvp.md for the build plan.
License
MIT — see LICENSE.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。