CG Alpha MCP

CG Alpha MCP

Enables crypto market analysis by combining Elfa's sentiment data and trending tokens with technical analysis tools (RSI, Bollinger Bands). Works alongside CoinGecko MCP to provide comprehensive token performance insights across multiple timeframes.

Category
访问服务器

README

cg-alpha-mcp

An MCP server that connects Elfa (crypto news & twitter sentiment data) and works alongside the CoinGecko MCP. Includes a tiny TA module (RSI + Bollinger Bands) to quickly inform users of token performance in various timeframes. Ideally used with CoinGecko feature top gainers/losers, which can be called directly from the LLM with CoinGecko Pro. Users can also connect the Nansen MCP for additional context in their token alpha searches.


Features

  • ELFA integrations
    • /v2/aggregations/trending-tokens
    • /v2/data/token-news
    • /v2/data/keyword-mentions
  • Generic ELFA proxy: call any ELFA path with elfa_query
  • Auth from .env (no key pasting into the chat)
  • TA utilities: RSI (Wilder) and Bollinger Bands (SMA + population stdev)
  • Designed to run with any MCP client (Claude Desktop, Cursor, etc. - n.b. designed with Claude desktop)

Prompts and responses

See the full set here: promts_responses_images/

Quick Start

1) Requirements

  • Node.js 20+ (works on 22 as well)
  • A modern MCP client (e.g., Claude Desktop or Cursor)

2) Install

git clone https://github.com/<you>/cg-alpha-mcp.git
cd cg-alpha-mcp
npm install   # installs dependencies (node_modules is ignored in Git)

3) Configure environment

cp .env.example .env

  • Edit .env and set ELFA_API_KEY (and COINGECKO_API_KEY if you use the CoinGecko MCP with a key)
  • Keep .env private — it’s already in .gitignore.

4) Hook into your MCP client

Most MCP clients let you add a custom server command. Use:

  • Command: node
  • Args: C:/Users/YOUR-FILE-NAME/cg-alpha-mcp/mcp-server.js (or ./mcp-server.js on macOS/Linux)
  • Working directory: the repo root On Windows, paths often look like: C:\Users\YOUR-FILE-NAME\cg-alpha-mcp\mcp-server.js

5) Once added, your client should show tools like:

  • elfa_status
  • elfa_reload_env
  • elfa_trending / elfa_trending_tokens
  • elfa_token_news
  • elfa_keyword_mentions
  • elfa_query

6) Environment Variables

See .env.example for all options.

  • ELFA_API_KEY (required for ELFA calls)
  • Get from ELFA (https://www.elfa.ai/api).
  • COINGECKO_API_KEY - add to Claude config file.

Advanced (optional):

  • ELFA_HEADER: header name for auth (defaults to x-elfa-api-key)
  • ELFA_AUTH_TYPE: set to x-elfa-api-key (default) or bearer
  • ELFA_BASE: ELFA base URL (defaults to https://api.elfa.ai)

7) Claude config file - structure at bottom of README

8) (Optional) - Nansen paid users can download .dxt file to install

  • Either drag and drop into Extensions or
  • Extensions > Advanced Setting > Install Extension > Choose .dxt file > Enter your API key

9) Troubleshooting:

401 “API key is required”

  • Check .env has ELFA_API_KEY=...
  • Run elfa_reload_env (no restart needed)
  • Verify header style: ELFA expects x-elfa-api-key (the server auto-tries both)

404 “Cannot GET …/data/trending”

  • Use /v2/aggregations/trending-tokens (the server’s elfa_trending already points here)
  • .env in the right place?
  • Put .env in the same folder as mcp-server.js (repo root)

Windows path issues

  • Use absolute paths in the MCP client if needed, e.g. C:\Users\YOUR-FILE-NAME\cg-alpha-mcp\mcp-server.js

Security

  • .env is ignored by git.
  • Include a public, safe .env.example so others can configure their own keys.

Useful Tools & What They Do

  • elfa_status - Shows current base URL and masked auth status.
  • elfa_reload_env - Reloads .env at runtime (no restart). Use this after editing the .env.
  • elfa_set_auth - Manually sets the API key and header style if needed: { "key": "sk-...", "headerName": "x-elfa-api-key", "scheme": "" }
  • elfa_trending / elfa_trending_tokens - Wrapper around /v2/aggregations/trending-tokens: { "timeframe": "24h", "limit": 10, "chain": "all" }
  • elfa_token_news - Token news: { "symbols": "BTC,ETH", "limit": 20, "start": "2025-08-01", "end": "2025-08-16" }
  • elfa_keyword_mentions - Multi-keyword mentions: { "keywords": ["bitcoin","halving"], "limit": 50 }
  • elfa_query - Generic ELFA proxy for any path: { "path": "/v2/aggregations/trending-tokens", "method": "GET", "query": { "timeframe": "24h", "limit": 10 } }

Technical Analysis (TA)

We provide a tiny pure JS TA module in services/ta.js:

  • rsi(values, period=14) → last RSI value (0–100)
  • bollinger(values, period=20, mult=2){ mean, upper, lower, last, percentB, bandwidth } Inputs are arrays of numeric closes in order oldest → newest. Your MCP orchestration fetches price series (e.g., via CoinGecko MCP) and then calls TA.

Intended Prompt Workflow:

"Show me the top 15 trending tokens on coingecko via their mcp, and the top 10 trending tokens on Elfa." (- Nansen users can ask for the recent smart money flows and activity.) "Show me the top 10 gainers and top 10 losers over the past 7 days in USD (specify timeframe: 24hours, 7days, 30days, specify token range: top 1000 or all tokens)." (- Goingecko Pro API users can call this from the MCP. Demo API users can ask Claude to determine this manually - *tokens with 24hr volume over 50k.)

User Chooses Specific Coins/Tokens of Interest from above lists:

"For (selected coins/tokens), show me the token news from Elfa AI, the trends in mentions and mentions delta from Elfa, and give me the RSI and Bollinger Bands for each." (- Nansen users can also ask for smart money data/comparison for these tokens - who is buying (selected coins/tokens)?)

Other prompts:

Built-in timeframe (fast): "Show me the top 10 gainers and top 10 losers over the past 7 days in USD, and give me the RSI and Bollinger Bands for each." Custom timeframe (computed): "Find the best and worst performing coins over the past 12 hours, limited to the top 200 by market cap. Add RSI and Bollinger Bands." Specific coins/tokens: "For BTC, ETH, and SOL, calculate the percentage change over the last 72 hours and display RSI(14) and Bollinger Bands(20,2)." Multi-window snapshot: "Create a snapshot: "top 5 gainers and losers for 24h, 7d, and 30d, with RSI and Bollinger for each result." Latest Trending Tokens (ELFA) "Show me the top trending tokens from ELFA." RSI Calculation "Fetch Bitcoin's last 50 closing prices from CoinGecko and calculate RSI." Bollinger Bands "Get ETH’s last 30 days of prices and calculate Bollinger Bands." Gainers & Losers "Find the top 5 gainers and losers in the last 24h and calculate their RSI."


Claude config file structure:

{
  "mcpServers": {
    "cg-alpha-mcp": {
      "command": "node",
      "args": [
        "/absolute/path/to/cg-alpha-mcp/mcp-server.js"
      ]
    },
    "coingecko_mcp_local": {
      "command": "npx",
      "args": [
        "-y",
        "@coingecko/coingecko-mcp"
      ],
      "env": {
        "COINGECKO_DEMO_API_KEY": "YOUR_DEMO_KEY",
        "COINGECKO_ENVIRONMENT": "demo"
      }
    }
  }
}

推荐服务器

Baidu Map

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。

官方
精选
JavaScript
Playwright MCP Server

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。

官方
精选
TypeScript
Magic Component Platform (MCP)

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。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

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。

官方
精选
本地
TypeScript
VeyraX

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。

官方
精选
本地
graphlit-mcp-server

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。

官方
精选
TypeScript
Kagi MCP Server

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。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

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。

官方
精选