How do I install Agent Trust?

Copy the stdio config for Cursor, Claude Desktop, Claude Code, or VS Code on this page. It uses the publisher's install, not a PlugMCP proxy.

Does Agent Trust need an API key?

The publisher lists these environment variables: EVM_PRIVATE_KEY, RESOURCE_SERVER_URL.

What transport does Agent Trust use?

Agent Trust runs locally over stdio.

Agent Trust

Q: When was Agent Trust last checked?

Last checked 2026-10-16. ERC-8004 agent trust scores via x402: free preview + paid USDC-on-Base signed scores (5 chains)

by io.github.Nikoble1926

ERC-8004 agent trust scores via x402: free preview + paid USDC-on-Base signed scores (5 chains)

Power your agent

Add this server

The publisher's own command. Not a PlugMCP endpoint.

Local command

npx -y @nikosble1926/[email protected]

Credentials from the publisher

EVM_PRIVATE_KEY — Optional. Base wallet private key (0x...) for a low-balance hot wallet with a little USDC. Required only for the paid tool; the free get_trust_preview works without it. Never use your main wallet.
RESOURCE_SERVER_URL — Optional. Override the trust API base URL. Defaults to https://trust.nsgoods.org.

Provider

Claude Code

EVM_PRIVATE_KEY=<secret> RESOURCE_SERVER_URL=<value> claude mcp add io.github.nikoble1926-agent-trust-mcp -- npx -y @nikosble1926/[email protected]

Run in a terminal.

Claude Desktop

{
  "mcpServers": {
    "io.github.nikoble1926-agent-trust-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@nikosble1926/[email protected]"
      ],
      "env": {
        "EVM_PRIVATE_KEY": "<secret>",
        "RESOURCE_SERVER_URL": "<value>"
      }
    }
  }
}

claude_desktop_config.json

Step-by-step: add this in Claude

Remote URL or a command on your machine? How the two installs differ

Run on PlugMCP

Host a copy with our keys and logs.

A new server you create. It is not this listing's official endpoint. Use the publisher install when you want their URL and their credentials.

Launch

Related MCP

Related skills

View all

agent-trust-oracle

Read only, pay per call trust scores for ERC-8004 agents, served over the x402 protocol. Every response is a signed, verifiable verdict.

What this is (for AI agents): a pay per call HTTP 402 endpoint that returns a signed, machine readable trust score for an ERC-8004 agent. No accounts, no API keys. Pay USDC on Base, get a verdict you can verify yourself.

What it does

Returns a signed trust score for an agent, on demand, per call over HTTP 402
Read only on chain analytics across multiple EVM chains (Ethereum, Base, BSC, Mantle, Polygon)
Built for the ERC-8004 agent identity and reputation ecosystem
Free preview endpoint so you can see the shape before you pay

How it works

Signed with EIP-191; every verdict is canonicalized and independently verifiable
Settled per call in USDC on Base
You pay only for a successful 2xx response
Open source, MIT

Live: trust.nsgoods.org Manifest: /.well-known/x402 · Methodology: /methodology

Part of nsgoods, a suite of signed x402 data oracles. Hub: x402.nsgoods.org