MCP.Market

web3agent MCP server

by Apegurus·io.github.Apegurus/web3agent·v0.6.2·1 stars
No reviews yet

Web3 MCP proxy for AI agents: DeFi swaps, bridges, market data, wallets, and gated writes.

B83/100grade B
Write a reviewInstall
What users say
No reviews yet
Be the first
Safety scan
B83/100

full report

Adoption
Growing

1 stars68 downloads/wk

Reviews

Write one

Nobody has reviewed web3agent yet.

If you have run it, two minutes of your experience saves the next person an afternoon.

web3agent tools (5)

write = sends, deletes, buys or posts

Read from the package source without running it. The installed server may list more.

  • explorer_get_daily_block_rewards

    Get daily block rewards for a date range.

  • explorer_get_daily_gas_used

    Get daily average gas used for a date range.

  • explorer_get_daily_new_addresses

    Get daily new address count for a date range.

  • explorer_get_daily_tx_count

    Get daily transaction count for a date range.

  • explorer_get_network_utilization

    Get daily network utilization percentage for a date range.

Public scan report

scanner v0.1.9 · 2026-09-26 · same rubric, same numbers if you re-run it

no findings
  • Code scan49 source files scanned25/25
  • –Live reliabilityno gateway calls yet and no remote to proben/a
  • –Tool poisoningtools not inspected (local package is not executed); not countedn/a
  • Auth qualitystatic API keys via environment variables6/15
  • Maintenancelast push 1 days ago15/15
  • Maintainer identityregistry namespace matches repository owner; GitHub account older than a year8/10
Overall 83/100. Components that don't apply are left out of the denominator. Any critical finding is an F.RubricAppeal a findingJSON

How web3agent compares

Against every graded MCP server, from our own scans and uptime checks. Updated after each scan run.

Safety score
83/100
higher than 65% of 48,255 servers
Tools
5
the typical server has 7
Tools that change things
0%
typical: 0%. Fewer write tools means less an agent can break
GitHub stars
1
more than 65% of 28,688 servers
Downloads a week
68
more than 53% of 12,191 servers with a package

What its tools let an agent do: 5 read

Install web3agent in Claude Code, Cursor or VS Code

Runs npx -y web3agent on your machine. Read the scan report first; the gateway never runs local packages.

claude mcp add web3agent -- npx -y web3agent
Add to Cursor

What the publisher says

From the web3agent repository's README, as published. We do not edit it. Read it on GitHub

See it in production: The Arena — 11 AI agents trade real capital through Web3Agent.

web3agent

MCP package: web3agent | npm: | GitHub: | Contact:

Links: Website · GitHub · npm · Smithery · The Arena · X / @Web3AgentFi

Give your AI agent EVM execution and DeFi tooling: swaps, bridges, limit and trigger orders, exchange trading, market data, research, wallet management. 190+ MCP tools. One install.

Works out of the box with Claude Code, Cursor, Windsurf, OpenCode, and Codex. Self-custodial. By default, write operations go through a confirmation queue: nothing executes without your approval unless you explicitly disable confirmations.

EVM execution is a solved problem. Stop rebuilding it. Plug in and ship.

What you can do

Once installed, your AI agent can execute real DeFi operations in plain language:

  • "Swap 0.1 ETH for USDC on Base" — quoted, routed, confirmation-gated, executed
  • "Bridge 500 USDC from Arbitrum to Optimism" — cross-chain via LI.FI, 20+ chains
  • "Set a limit or trigger order to buy ETH at $2,800" — decentralized Spot orders via Orbs
  • "Cancel my open orders on Binance" — CCXT exchange access with per-method risk classification
  • "What's my USDC balance on Base?" — read-only EVM and ERC-20 balance checks
  • "Show me yield opportunities above 5% APY" — research tools, protocol analysis, due diligence

Common flows avoid ABI handling and transaction building. Generic contract reads and writes are available when you provide or register an ABI, or when explorer ABI lookup is available. The agent handles routing, transaction preparation, and the confirmation queue. You approve, it executes.

Quickstart for agent power users

npx web3agent init

This detects Claude Code, Cursor, Windsurf, OpenCode, or Codex and configures the MCP server automatically. Restart your host, then ask:

What Web3 tools do you have available, and what chain are you configured for?

For any other MCP-capable host, configure a stdio server that runs:

npx web3agent

The generic MCP config shape is:

{
  "web3agent": {
    "type": "stdio",
    "command": "npx",
    "args": ["web3agent"]
  }
}

Writes are confirmation-gated by default, and wallet secrets are not exposed through MCP unless explicitly enabled.

For a step-by-step guide covering both human and agent setups, see docs/guides/universal-access.md.

Smithery local bundle

Smithery URL publishing requires a hosted Streamable HTTP MCP endpoint. Web3Agent is distributed as a local stdio/npm server, so Smithery distribution uses an MCPB bundle instead:

pnpm run mcpb:check
smithery mcp publish dist/web3agent.mcpb -n Apegurus/web3agent

The generated MCPB is a thin local bundle that launches the published npm package with npm exec --package web3agent@0.6.2.

Why Web3Agent

Shortened. The full README is on GitHub.

Nothing above is checked by us. What we check is on the safety report.

web3agent: common questions

Is web3agent MCP server safe?
Mostly: it is graded B (83/100). Read the web3agent safety report
How do I install web3agent?
It runs on your machine. Copy the Claude Code, Cursor, VS Code or Claude Desktop config from the install section.
Does web3agent need an API key?
Yes. The registry entry asks for ETHERSCAN_API_KEY, ZEROX_API_KEY, COINGECKO_API_KEY, OWS_PASSPHRASE.
Is web3agent maintained?
The last commit was 40 days ago (2026-09-25). The latest release is v0.6.2.
How does web3agent compare with other MCP servers?
safety score 83/100 (higher than 65% of 48,255 servers); tools 5 (the typical server has 7); tools that change things 0% (typical: 0%. Fewer write tools means less an agent can break). Full comparison.