MCP Catalogs
← Home

tinymcp

by golioth·★ 150·Score 43

MCP server enabling LLMs to control embedded devices via Golioth API and RPCs.

iotdeveloper-toolsai-llm
5
Forks
3
Open issues
11 mo ago
Last commit
2d ago
Indexed

Overview

tinymcp is an MCP server that bridges LLMs with embedded devices using the Golioth IoT platform. It exposes device functionality through LightDB State and RPCs, allowing AI systems to interact with physical hardware without modifying device firmware. The server runs locally and communicates with remote devices via the Golioth management API, making it possible to delegate physical capabilities to AI systems while maintaining security through proper authentication.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

you:LLM-controlled IoT device automation and management
you:Remote device monitoring and control through natural language
you:Physical interaction with devices via AI agents
you:Is tinymcp suitable for production use?
you:What MCP clients are compatible with tinymcp?

When to choose this

Choose tinymcp when you need AI systems to interact with physical embedded devices and you're already using or willing to adopt the Golioth platform.

When NOT to choose this

Avoid tinymcp if you don't want vendor lock-in to Golioth, need direct device control without an intermediary platform, or require production stability as it's still experimental.

Comparable tools

embedded-llm-serverdevice-mcpiot-mcp-server

Installation

Installation

Prerequisites

  • Go installed
  • Golioth account and project
  • Connected device

Steps

  1. Set environment variables:
export TINYMCP_PROJECT=<your-golioth-project-id>
export TINYMCP_DEVICE=<your-golioth-device-id>
export TINYMCP_API_KEY=<api-key-for-your-golioth-project>
  1. Build and run the server:
go build -o tinymcp ./server
./tinymcp
# or directly:
go run ./server
  1. For testing with MCP Inspector, connect to http://localhost:8080

FAQ

Is tinymcp suitable for production use?
No, tinymcp is currently an experimental project with frequent breaking changes. Extreme caution should be taken when delegating physical capabilities to AI systems.
What MCP clients are compatible with tinymcp?
Most MCP client implementations should be compatible. Documentation specifically mentions MCP Inspector, Claude Code, and Gemini CLI as tested options.

On Hacker News

Recent discussion from the developer community.

Compare tinymcp with

tinymcp vs ultimate_mcp_servertinymcp vs mcp-server-charttinymcp vs everythingtinymcp vs filesystemtinymcp vs time
GitHub →

Last updated · Auto-generated from public README + GitHub signals.