AgentScore MCP Server

1

Add it to Claude Code

Run this in a terminal.

Run in terminal
claude mcp add -e "AGENTSCORE_API_KEY=${AGENTSCORE_API_KEY}" agentscore -- npx -y @agent-score/mcp
Required:AGENTSCORE_API_KEY+ 1 optional
README.md

Trust and reputation tools for AI agent wallets

@agent-score/mcp

Model Context Protocol server for AgentScore — trust and reputation tools for AI agent wallets in the x402 payment ecosystem and ERC-8004 agent registry.

Tools

Tool Tier Description
check_wallet_reputation Free Cached trust reputation lookup
assess_wallet Paid On-the-fly assessment with allow/deny decision
browse_agents Free Browse registered ERC-8004 agents
get_ecosystem_stats Free Ecosystem-level statistics

Setup

Get an API key at agentscore.sh/sign-up.

Claude Code

claude mcp add agentscore -e AGENTSCORE_API_KEY=ask_... -- npx -y @agent-score/mcp

Cursor

Add to .cursor/mcp.json:

{
  "mcpServers": {
    "agentscore": {
      "command": "npx",
      "args": ["-y", "@agent-score/mcp"],
      "env": {
        "AGENTSCORE_API_KEY": "ask_..."
      }
    }
  }
}

Claude Desktop

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "agentscore": {
      "command": "npx",
      "args": ["-y", "@agent-score/mcp"],
      "env": {
        "AGENTSCORE_API_KEY": "ask_..."
      }
    }
  }
}

Configuration

Variable Required Default Description
AGENTSCORE_API_KEY Yes API key from agentscore.sh
AGENTSCORE_BASE_URL No https://api.agentscore.sh API base URL override

Documentation

License

MIT

Tools (4)

check_wallet_reputationPerforms a cached trust reputation lookup for a given wallet address.
assess_walletProvides an on-the-fly risk assessment with an allow or deny decision.
browse_agentsAllows browsing of registered ERC-8004 agents.
get_ecosystem_statsRetrieves ecosystem-level statistics for the AgentScore platform.

Environment Variables

AGENTSCORE_API_KEYrequiredAPI key from agentscore.sh
AGENTSCORE_BASE_URLAPI base URL override

Configuration

claude_desktop_config.json
{"mcpServers": {"agentscore": {"command": "npx", "args": ["-y", "@agent-score/mcp"], "env": {"AGENTSCORE_API_KEY": "ask_..."}}}}

Try it

Check the reputation of wallet address 0x123... to see if it is trusted.
Assess the risk of interacting with this agent wallet address.
Browse the current list of registered ERC-8004 agents.
What are the current ecosystem statistics for AgentScore?

Frequently Asked Questions

What are the key features of AgentScore?

Cached trust reputation lookups for AI agent wallets. Real-time risk assessment with allow/deny decisioning. Integration with ERC-8004 agent registry. Support for x402 payment ecosystem protocols.

What can I use AgentScore for?

Verifying the trustworthiness of an AI agent before authorizing a transaction. Filtering out malicious or low-reputation agents in an automated workflow. Discovering verified agents within the ERC-8004 registry. Monitoring ecosystem-wide trends in AI agent wallet security.

How do I install AgentScore?

Install AgentScore by running: Add to claude_desktop_config.json

What MCP clients work with AgentScore?

AgentScore works with any MCP-compatible client including Claude Desktop, Claude Code, Cursor, and other editors with MCP support.

Turn this server into reusable context

Keep AgentScore docs, env vars, and workflow notes in Conare so your agent carries them across sessions.

Need the old visual installer? Open Conare IDE.
Open Conare