Meta Open-Sources Muse SDKs for DIY Gadgets

Meta is opening two routes for Muse to interact with hardware: programmable gadgets built with its open-source SDKs, and a Home Link device that connects the AI agent to compatible equipment on a home network.

DIY builders can adapt the released code for ESP32 boards or Raspberry Pi/Linux systems, adding displays, buttons, audio, sensors, and controls. Home Link offers a ready-made network bridge, but it runs only official firmware and cannot be reflashed.

The Muse Gadgets project gives developers a way to extend Muse beyond its app. It also raises practical questions about permissions, device compatibility, and community-written instructions that can trigger actions on a local network. Before building or connecting anything, it helps to separate what Meta provides from what remains the builder’s responsibility.

What Builders Can Make With the Muse SDKs

Meta’s release includes an ESP32 Device SDK for microcontroller-based projects and a Linux Device SDK for Raspberry Pi or other Linux systems.

The ESP32 route supports projects with screens, audio input and output, and other connected sensors. Meta’s examples include a small status display, a push-to-talk device with a microphone and speaker, and a color e-ink display for briefings, reminders, or shopping lists.

These are project ideas and starting points, not finished Meta products. The site explicitly says the featured third-party devices are made and sold by other companies and that Meta does not endorse or warrant them. The TV-connected HDMI stick is also marked “coming soon,” so its appearance alongside existing hardware examples should not be read as availability.

With the Linux SDK, builders can add custom commands so Muse can handle system-administration tasks or work with a Home Assistant setup. Meta describes these uses as options for customization. A spare Raspberry Pi could serve as both an interface and a host for custom actions, though which actions work depends on what the builder implements.

The official Muse Gadget SDK repository contains the ESP32 and Linux code, firmware, and a skills directory. The main release uses Apache-2.0, with licensing exceptions for some third-party files, dependencies, and assets. Anyone redistributing a build should check those exceptions before assuming the entire project has the same license.

Meta provides the SDKs and firmware as-is, without warranty. Builders can inspect and modify the open-source device-side code, but Muse itself is not open source, and connecting to the service remains subject to separate conditions.

How ESP32 and Linux Gadgets Pair With Muse

The repository documents a shared pairing process for DIY ESP32 and Linux gadgets:

1. Obtain an SDK token and review the Gadget SDK Terms. Every gadget needs a token to pair.
2. Follow the relevant SDK directory’s README. The ESP32 and Linux directories contain their own getting-started instructions.
3. Open the Muse app on iOS or Android. Go to Settings > Devices and enable Developer mode.
4. Look for a device prefixed with “MuseGadget” and pair it through the app.

Builders need working device software as well as these account-side pairing prerequisites. Apache-2.0 governs the released code; access to Muse requires following the SDK’s documented pairing process and terms.

Treat pairing tokens as credentials: don’t publish them in project repositories, screenshots, or shared configuration files. This is a practical precaution, not a claim that the release has a token-handling vulnerability.

Home Link Reaches Devices Through Local HTTP APIs

Muse Home Link gives Muse a route to compatible equipment already on a home network. Users plug in a small device instead of building a new interface.

Meta describes Home Link as an ESP32-C5 gadget with USB-C and USB-A connectivity. Its listed hardware includes Wi-Fi 6 on both 2.4 GHz and 5 GHz networks, plus an LED for pairing and device status.

Setup starts with a USB power source near the router. Users add Home Link in the Muse app, complete setup over Bluetooth LE, and choose the Wi-Fi network it should join. Home Link stays on that network so Muse can reach compatible devices through local HTTP APIs.

A local HTTP API is an interface a device or service exposes over the network. Wi-Fi connectivity alone is not enough: the target must provide an API that the relevant skill can use. Meta’s examples include switching on a light, controlling a TV, or sending a document to a printer. Each depends on the device, its configuration, and the available community skill.

Home Link’s firmware derives from the open-source ESP32 SDK, but Home Link itself cannot be reflashed. It runs only official firmware. Builders who want to replace or substantially customize the firmware need to use their own ESP32 hardware.

That restriction limits experimentation on the supplied unit. It does not establish that every action reachable through Home Link is safe.

Community Skills Are Not Official Device Integrations

Meta’s Home Link page points users toward community skills for products including Philips Hue, Sonos, Apple TV, Google Nest speakers, and Samsung TVs.

These are potential community-supported connections, with no guarantee that every model or configuration will work. Meta warns that skills can change or break and says it does not endorse or warrant the third-party devices mentioned.

There are three separate pieces to evaluate:

• The SDKs and firmware: code Meta publishes for building Muse gadgets.
• Home Link: hardware Meta offers that runs official firmware.
• Community skills: device-specific behavior contributed by others.

An official Home Link device does not turn an installed community skill into a Meta-supported integration. A project’s appearance on the Muse Gadgets site also does not establish that its hardware manufacturer supports connecting it to Muse.

Before buying hardware around a skill, check whether it supports your exact device and required actions. Review its documentation, maintenance history, and configuration requirements.

Meta explicitly advises against relying on community skills for safety-critical uses, including home security, emergencies, or medical needs.

Before You Connect It: A Security Checklist

An AI request can now lead to a network request or a custom Linux command, with effects outside the chat window. These precautions are deployment advice, not findings of a specific vulnerability in Muse or Home Link.

• Read the skill before installing it. Inspect the instructions, scripts, dependencies, and configuration. Identify the devices it contacts, the credentials it uses, and the actions it can trigger. If you cannot assess that access, start with a narrowly scoped example instead of an unfamiliar automation.

• Expose only the commands you need. For Linux projects, prefer specific actions over unrestricted shell execution. A command to retrieve system status has a different risk profile from one that can install software, delete files, or change permissions.

• Check endpoint authentication. Establish whether each local HTTP API requires credentials and which operations those credentials permit. A device being reachable only on your home network is not a substitute for appropriate authentication.