Repolode
Submit
aaronjmars

aaronjmars/web3-research-mcp

A local MCP server that gives AI assistants tools for researching crypto tokens using CoinGecko, DeFiLlama and web search, with no API key required.

View on GitHubWebsite
Stars
171
Forks
55
Open issues
0
Last commit
2 days ago

What is it?

Web3 Research MCP is a Model Context Protocol server that runs locally and lets MCP-compatible clients such as Claude Desktop, Claude Code and Cursor research a crypto token. It can create a structured research plan, track the status of each section, run web searches, and fetch pages and save them as resources.

For market and protocol data, it calls the public CoinGecko and DeFiLlama APIs directly rather than scraping HTML. CoinGecko lookups cover price, market cap, price changes, supply, contract addresses and exchange listings, while DeFiLlama lookups cover TVL, fees and per-chain breakdowns. An optional CoinGecko Pro key can be set through an environment variable.

The server is published on npm and started with npx, so no install step is needed. Search and data requests go to external providers, while research state and saved resources stay on the local machine.

Key features

  • 01Creates a structured research plan for a token and tracks progress by section.
  • 02Pulls live CoinGecko market data, exchange listings and contract addresses without an API key.
  • 03Retrieves DeFiLlama protocol TVL, fees and per-chain breakdowns.
  • 04Runs web and keyword searches and saves results as resources.
  • 05Fetches a URL and converts the content to text, HTML, markdown or JSON.

Use cases

  • Getting a quick market overview of a token from CoinGecko data inside an AI assistant.
  • Checking where a token trades and how liquid each venue is.
  • Comparing a DeFi protocol's TVL trend and fee income against its peers.
  • Building a sourced research report on a crypto project with saved web references.

Quick start

  1. 1

    Run the server with npx (no install or API key needed)

    npx -y web3-research-mcp@latest
  2. 2

    Add the server to Claude Code

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

Things to know

  • Some websites block scraping, so direct page fetches may fail with 403 errors.
  • Search results may not be comprehensive, and rate limits may apply to searches.

In collections

MCP servers for Claude CodeClaude Code tools, skills & pluginsOpen-source AI agent frameworks & appsMCP servers

Topics

#mcp#crypto#web3#ai-agent#defi#market-data#claude#coingecko#defillama#mcp-crypto

FAQ

What is web3-research-mcp?
A local MCP server that gives AI assistants tools for researching crypto tokens using CoinGecko, DeFiLlama and web search, with no API key required. Web3 Research MCP is a Model Context Protocol server that runs locally and lets MCP-compatible clients such as Claude Desktop, Claude Code and Cursor research a crypto token. It can create a structured research plan, track the status of each section, run web searches, and fetch pages and save them as resources.
Is web3-research-mcp free to use? What license does it have?
web3-research-mcp is open source under the MIT license. Check the license terms before using it in commercial products.
Is web3-research-mcp still maintained?
Actively maintained. The last commit to web3-research-mcp was on Nov 5, 2026.
What can I use web3-research-mcp for?
Getting a quick market overview of a token from CoinGecko data inside an AI assistant. Checking where a token trades and how liquid each venue is. Comparing a DeFi protocol's TVL trend and fee income against its peers. Building a sourced research report on a crypto project with saved web references.
How do I install web3-research-mcp?
Run the server with npx (no install or API key needed): npx -y web3-research-mcp@latest Add the server to Claude Code: claude mcp add web3-research-mcp -- npx -y web3-research-mcp@latest
Where can I get web3-research-mcp?
The source code and installation instructions are on GitHub: https://github.com/aaronjmars/web3-research-mcp

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects