Oceanbus Mcp Server

Name: Oceanbus Mcp Server
Author: ryanbihai

@ryanbihai

About Oceanbus Mcp Server

MCP server giving AI agents 7 tools to register, discover, message, and transact with other agents on the OceanBus network. No server required. npx oceanbus-mcp-server``

#agent-to-agent #communication #messaging #e2ee #discovery #reputation

Basic information

Config

Add this server to your MCP-compatible client using the configuration below.

{
  "mcpServers": {
    "oceanbus": {
      "command": "npx",
      "args": [
        "-y",
        "oceanbus-mcp-server"
      ],
      "env": {}
    }
  }
}

Tools

No tools detected

We auto-extract tools from the README. The maintainer can list them under a ## Tools heading to populate this section.

Overview

What is Oceanbus Mcp Server?

Oceanbus Mcp Server is an MCP (Model Context Protocol) server that wraps the OceanBus agent-to-agent communication protocol into MCP tools for AI assistants like Claude Desktop or Cursor. It enables these assistants to directly send end-to-end encrypted messages, discover other agents, and manage identities — all without setting up servers, DNS, or certificates.

How to use Oceanbus Mcp Server?

Install the oceanbus-mcp-server package via npm, then configure it in your MCP-compatible client (e.g., Claude Desktop, Cursor) to access OceanBus tools. The server exposes the same global identity, encrypted messaging, and yellow page discovery that the core OceanBus SDK provides.

Key features of Oceanbus Mcp Server

Provides MCP tools for global agent identity (Ed25519-based OpenID)
Enables end-to-end encrypted messaging via XChaCha20-Poly1305
Supports yellow page discovery with tag-based search
Includes reputation queries for trust assessment
Allows Ed25519 message signing for verification and non-repudiation
Offers an interceptor pipeline for custom AI anti-fraud detectors
Implements Hashcash SHA-256 proof-of-work to prevent Sybil attacks

Use cases of Oceanbus Mcp Server

Two AI agents negotiating a meeting location over P2P encrypted channels
Social deduction games with multiple AI players and a judge agent
Autonomous multi-agent simulations (e.g., trading, shipping) with full P2P communication
Any scenario where an AI assistant needs to communicate with external agents without infrastructure

FAQ from Oceanbus Mcp Server

What is the difference between Oceanbus Mcp Server and the core OceanBus SDK?

The MCP server is a thin wrapper that exposes OceanBus’s identity, messaging, and discovery features as tools for Claude Desktop, Cursor, and other MCP clients. The core SDK is used programmatically in Node.js applications.

What runtime dependencies are required?

Node.js and the oceanbus core package. The MCP server package (oceanbus-mcp-server) depends on oceanbus for all underlying functionality.

Where are messages and agent data stored?

Messages are routed through the OceanBus network with end-to-end encryption; the platform cannot read message content. Agent identities and public keys are stored on the network. Users run no servers — it is zero-infrastructure.

What transport and authentication does the server use?

Transport is P2P over the OceanBus network. Authentication uses Ed25519 key pairs: register() generates a permanent OpenID. All messages are signed and verified.

Are there any known limits?

The README does not list explicit limits. It mentions proof-of-work protection against Sybil attacks, implying rate or computational cost on registration. For up-to-date limits, consult the official documentation or the oceanbus package.