NATS
MCP server for NATS — publish, request-reply, JetStream streams, KV store, and server diagnostics over stdio
nats-mcp
MCP server for NATS — publish messages, request-reply, manage JetStream streams, use the KV store, and inspect server diagnostics over stdio.
Installation
bunx @daanrongen/nats-mcp
Tools (17 total)
| Domain | Tools | Coverage |
|---|---|---|
| Core | publish, request | Publish messages, request-reply pattern |
| Streams | stream_list, stream_info, stream_create, stream_delete, stream_publish, stream_fetch, stream_consumer_create | JetStream stream lifecycle and message delivery |
| KV | kv_create_bucket, kv_list_buckets, kv_get, kv_put, kv_delete, kv_list_keys, kv_history | Key-value store with history |
| Server | server_info | NATS server diagnostics |
Configuration
| Variable | Required | Description |
|---|---|---|
NATS_URL | Yes | NATS server URL (e.g. nats://localhost:4222) |
Setup
Claude Desktop
Edit ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"nats": {
"type": "stdio",
"command": "bunx",
"args": ["@daanrongen/nats-mcp"],
"env": {
"NATS_URL": "nats://localhost:4222"
}
}
}
}
Claude Code CLI
claude mcp add nats -e NATS_URL=nats://localhost:4222 -- bunx @daanrongen/nats-mcp
Development
bun install
bun run dev # run with --watch
bun test # run test suite
bun run lint # lint and check formatting
bun run typecheck # type check
bun run build # bundle to dist/main.js
bun run inspect # open MCP Inspector in browser
Inspecting locally
bun run inspect launches the MCP Inspector against the local build:
bun run build && bun run inspect
This opens the Inspector UI in your browser where you can call any tool interactively and inspect request/response shapes.
Architecture
src/
├── config.ts # Effect Config — NATS_URL
├── main.ts # Entry point — ManagedRuntime + StdioServerTransport
├── domain/
│ ├── NatsClient.ts # Context.Tag service interface
│ ├── errors.ts # NatsError, KvNotFoundError, StreamNotFoundError
│ └── models.ts # Schema.Class models (NatsMessage, StreamInfo, KvEntry, …)
├── infra/
│ ├── NatsClientLive.ts # Layer.scoped — connects via nats.js, drains on shutdown
│ └── NatsClientTest.ts # In-memory Ref-based test adapter
└── mcp/
├── server.ts # McpServer wired to ManagedRuntime
├── utils.ts # formatSuccess, formatError
└── tools/ # core.ts, streams.ts, kv.ts, server.ts
License
MIT
Servidores relacionados
Africa's Talking Airtime MCP
Interact with Africa's Talking airtime service and store transaction data in a local SQLite database.
Prompt for User Input MCP Server
Enables AI models to prompt users for input directly within their code editor for interactive conversations.
nadanada_me
A public MCP server that gives AI agents access to real UK carrier phone numbers for SMS verification. Agents can rent disposable or rental numbers, pay Lightning invoices, and read incoming SMS, all through standard MCP tool calls with no authentication required.
Fastmail MCP Server
Integrate your Fastmail account with AI assistants using the Model Context Protocol.
TextForge
The MCP email server with human approval — AI drafts emails, you approve before they send. Works with Claude Code, Cursor, Windsurf, and any MCP client.
mcp-gmail
MCP server for full Gmail operations via Unipile API. 9 tools: send, reply, list, read, delete, search, labels, attachments, drafts. Dry-run by default, 55 unit tests. MIT licensed.
MCP Relay
A Discord relay server to send messages and prompts to a channel and receive responses.
WasenderAPI MCP
Connect WasenderAPI to AI agents and automation tools using the Model Context Protocol. This remote MCP server exposes WhatsApp session management, messaging, contacts, and groups as callable tools for MCP-compatible platforms
BAGO
BAGO — AI-first community where AI agents register, post, and govern
Unichat
A unified interface for various chat AI models including OpenAI, MistralAI, Anthropic, and Google AI, requiring vendor API keys.