dvmcp
DVMCP is a bridge implementation that connects Model Context Protocol (MCP) servers to Nostr's Data Vending Machine (DVM) ecosystem
claude mcp add --transport stdio gzuuus-dvmcp node packages/dvmcp-discovery/dist/index.js \ --env LOG_LEVEL="info" \ --env DVMCP_CONFIG_PATH="Path to config if needed (e.g., ./config/dvmcp-discovery.json)"
How to use
DVMCP is a bridge between MCP servers and Nostr's Data Vending Machine (DVM) ecosystem. It enables MCP-powered tools and capabilities to be discovered, requested, and executed via decentralized Nostr messages, using JSON-RPC style interactions. The discovery service aggregates tools declared by MCP servers and exposes them over the DVMCP protocol so clients on Nostr networks can discover what services are available, understand their capabilities, and invoke them in a verifiable, cryptographically signed manner. This makes MCP-based AI or computation services accessible without centralized registries while preserving interoperability with Nostr event kinds and data exchange flows.
To use it, run the DVMCP discovery service (and the bridge when needed). The bridge publishes tool announcements and status updates to DVMs, while the discovery service collects MCP tools from participating servers and makes them available to DVMs. Clients can query the tools list, resources, prompts, and tool capabilities, then submit requests and receive responses through the DVM workflow, all in a decentralized manner.
How to install
Prerequisites:
- A supported runtime (Node.js/Bun as used in this project).
- Git to clone the repository.
-
Clone the repository: git clone https://github.com/gzuuus/dvmcp.git cd dvmcp
-
Install dependencies (using Bun as described in the repository): bun install
-
Run the discovery service and bridge (quick start): bunx dvmcp-discovery bunx dvmcp-bridge
If you prefer a global style setup (per the project's quick start):
-
Global installation ( Bun-based): bun install -g @dvmcp/bridge @dvmcp/discovery
-
Run the commands directly: dvmcp-bridge dvmcp-discovery
Note: The project is archived and superseded by ContextVM/sdk, so consider upgrading to the newer ContextVM tooling for production use. The commands above reflect the repository’s Quick Start as provided in the README.
Additional notes
Tips and considerations:
- The project is archived and superseded by ContextVM/sdk; for production workloads, prefer the modern SDKs and tooling from that project.
- Prerequisite tool: Bun must be installed for the quick-start commands (bunx, bun, etc.).
- The DVMCP workflow relies on Nostr event kinds for server announcements, tool lists, resources, prompts, and request/response messages; ensure your DVM client can sign and verify these messages.
- Environment variables (as placeholders) can be used to configure logging, config paths, and operational modes. Customize as needed for your deployment.
- If you run into issues, check that the discovery and bridge services are accessible and that their published tool metadata (tools list, resources, prompts) are synchronized.
- This MCP server set supports decentralized discovery and cryptographic verification, but always validate tool configurations and authorization in your specific DVM/network context.
Related MCP Servers
markdownify
A Model Context Protocol server for converting almost anything to Markdown
mcp-chat
Examples of using Pipedream's MCP server in your app or AI agent.
toolbase
A desktop application that adds powerful tools to Claude and AI platforms
codemesh
The Self-Improving MCP Server - Agents write code to orchestrate multiple MCP servers with intelligent TypeScript execution and auto-augmentation
mcp-easy-installer
MCP easy installer is a robust mcp server with tools to search, install, configure, repair and uninstall MCP servers
docmole
Dig through any documentation with AI - MCP server for Claude, Cursor, and other AI assistants