mcpField
W
Web3 Research MCP

Web3 Research MCP

Local
github.com/aaronjmarsData & Analytics170Updated 6d ago

About Web3 Research MCP

Web3 Research MCP is a locally run server for researching cryptocurrency tokens, intended for crypto research workflows. It combines web search and direct CoinGecko and DeFiLlama data with saved resources and section status tracking.

Key features

  1. 01Create structured token research plans
  2. 02Search web, news, images, and videos
  3. 03Fetch CoinGecko market and exchange data
  4. 04Fetch DeFiLlama protocol data
  5. 05Save resources and track research status

Use cases

  1. 01Gathering market data for a token report
  2. 02Checking a token across search sources
  3. 03Comparing protocol TVL and fee data

MCP support

It exposes research, search, data-fetching, resource, and status tools, plus a token-research prompt, to MCP clients. Research state and saved resources are stored locally.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

mcp.json · Universal config
{
"mcpServers": {
"web3-research": {
"command": "npx",
"args": [
"-y",
"@smithery/cli"
]
}
}
}
Run commandnpx -y @smithery/cli

Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.

Packages
npm@smithery/cli

Tools

No tool list yet

Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.

Details

From the project README

Web3 Research MCP - deep research for crypto from a locally run server. Searches and data requests go to external providers; research data is stored locally.

Star us ❤️ →   Star Web3 Research MCP on GitHub   web3-research-mcp on npm   @aaronjmars on X

Deep crypto research from a locally run MCP server.
Searches and data requests go to their external providers, and page-fetch tools contact the supplied URLs. Research state and saved resources are stored locally; public sources need no API key.

stars forks npm license

## 🚀 Preview

Preview Preview2

🧠 Features

  • Comprehensive Research: Gather detailed information about any cryptocurrency token
  • Multi-Source Analysis: Research across multiple sources including CoinGecko, CoinMarketCap, DeFiLlama, and more
  • Structured Reporting: Generate detailed reports covering technical fundamentals, market data, social sentiment, and more
  • Resource Management: Automatically stores search results and content for reference
  • Status Tracking: Track research progress through different stages and sections

📋 Requirements

  • Node.js 18 or higher

🔧 Installation & Setup

Installing via Smithery

To install web3-research-mcp for Claude Desktop automatically via Smithery:

npx -y @smithery/cli install web3-research-mcp --client claude

🔌 Using with Claude Desktop

Edit your Claude Desktop config file

  • Windows: %APPDATA%\Claude\claude_desktop_config.json
  • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json

Add this to your Claude Desktop configuration file:

{
  "mcpServers": {
    "web3-research-mcp": {
      "command": "npx",
      "args": ["-y", "web3-research-mcp@latest"]
    }
  }
}

Then restart Claude Desktop

🔌 Using with Claude Code

claude mcp add web3-research-mcp -- npx -y web3-research-mcp@latest

🔌 Using with Cursor

Go to: Settings -> Cursor Settings -> MCP -> Add new global MCP server Paste this into your Cursor ~/.cursor/mcp.json file. See Cursor MCP docs for more info.

{
  "mcpServers": {
    "web3-research-mcp": {
      "command": "npx",
      "args": ["-y", "web3-research-mcp@latest"]
    }
  }
}

Then restart Cursor

🧩 Run from source

git clone https://github.com/aaronjmars/web3-research-mcp.git
cd web3-research-mcp
npm ci
npm run build
npm start   # MCP server over stdio

To point a client at your local build, use "command": "node" with "args": ["/path/to/web3-research-mcp/dist/server.js"].

⚙️ Configuration

No API key is needed. One optional env var:

  • COINGECKO_API_KEY: CoinGecko Pro key. When unset, the free public API is used.

🛠️ Tools

create-research-plan

Creates a structured research plan for a token.

Parameters:

  • tokenName: Full name of the token
  • tokenTicker: Ticker symbol of the token
search

Performs a web search and returns the results.

Parameters:

  • query: Search query
  • searchType: Type of search (web, news, images, videos)
research-with-keywords

Searches for a token with specific keywords and saves the results.

Parameters:

  • tokenName: Name of the token
  • tokenTicker: Ticker symbol
  • keywords: Array of keywords to search for
update-status

Updates the status of a research section.

Parameters:

  • section: Section name to update (e.g., 'projectInfo', 'technicalFundamentals')
  • status: New status for the section (planned, in_progress, completed)
fetch-content

Fetches content from a URL and saves it as a resource.

Parameters:

  • url: URL to fetch content from
  • format: Output format (text, html, markdown, json)
list-resources

Lists all available resources that have been saved.

search-source

Searches for information about a token from a specific source.

Parameters:

  • tokenName: Name of the token
  • tokenTicker: Ticker symbol
  • source: Source to search (e.g., 'CoinGecko', 'DeFiLlama', 'News')
research-source

Searches a specific source for a token and saves the top result's page content as a markdown resource.

Parameters:

  • tokenName: Name of the token
  • tokenTicker: Ticker symbol of the token
  • source: Source to research (e.g., 'IQ Wiki', 'CoinMarketCap')
coingecko-data

Fetches live market data directly from the CoinGecko public API: price, market cap, 24h/7d/30d changes, ATH/ATL, circulating supply, contract addresses across chains, and social/dev links. Bypasses the 403 issues of HTML scraping.

Parameters:

  • tokenName: Full name of the token (e.g., 'Bitcoin')
  • tokenTicker: Ticker symbol (e.g., 'BTC')

No API key required. Uses the free public tier (~30 req/min).

Optional: set COINGECKO_API_KEY in the environment to use a CoinGecko Pro API key. When set, requests are sent to https://pro-api.coingecko.com/api/v3 with the x-cg-pro-api-key header. Requests time out after 15s.

coingecko-search

Searches CoinGecko's coin index and returns candidate matches with their CoinGecko IDs. Useful when the ticker is ambiguous (e.g., multiple tokens with the same symbol).

Parameters:

  • query: Search query: name, ticker, or contract address
coingecko-tickers

Fetches active exchange listings for a token from CoinGecko: which CEXs/DEXs trade the pair, per-venue 24h USD volume, last price, bid-ask spread, trust score, and trade URL. Sorted by 24h USD volume desc; anomalies and stale prints are filtered out. Bypasses HTML scraping for the "where does this token actually trade" question.

Parameters:

  • tokenName: Full name of the token (e.g., 'Bitcoin')
  • tokenTicker: Ticker symbol (e.g., 'BTC')
  • limit: How many top venues to return (default 15, max 50)

No API key required. Uses the free public tier. Requests time out after 15s.

defillama-data

Fetches protocol data directly from the DeFiLlama public API: total TVL, ATH TVL with date, 30/90/365-day TVL change vs current, market cap (when available, since DeFiLlama omits it for many protocols), per-chain TVL breakdown, fees (24h/7d/30d/all-time), token addresses, fundraising rounds, and links. Bypasses HTML scraping for the most common DeFi-protocol lookup.

Parameters:

  • tokenName: Full protocol/token name (e.g., 'Uniswap')
  • tokenTicker: Ticker symbol (e.g., 'UNI')

No API key required. Uses the free public API. Requests time out after 15s. The protocol index is cached for 5 minutes per process to avoid hammering /protocols on every call.

defillama-search

Searches DeFiLlama's protocol index and returns candidate matches with their slugs, TVL, and category. Useful when the ticker is ambiguous (e.g., multiple protocols with similar names).

Parameters:

  • query: Search query: protocol name, ticker, or slug

📝 Prompts

token-research

Initiates comprehensive research on a cryptocurrency token.

Parameters:

  • tokenName: Full name of the cryptocurrency token
  • tokenTicker: Ticker symbol of the token (e.g., BTC, ETH)

🧠 How It Works

  1. When research begins, a structured plan is created covering all aspects of the token
  2. The server performs searches across multiple sources for information
  3. Search results are stored as resources that can be referenced
  4. The research progresses through different sections, with status tracking
  5. A comprehensive report is generated covering all aspects of the token

⚠️ Limitations

  • Some websites block web scraping, so direct content fetching may fail with 403 errors
  • Relies on search results which may not always be comprehensive
  • Rate limits may apply to search operations

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.


Built by Aaron Elijah Mars, founder of Aeon and MiroShark · @aaronjmars

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.