Building a DIY Muse AI Pendant for $50

The Waveshare ESP32-S3 1.75C is a compact touchscreen pendant officially sanctioned by Meta to run its Muse AI cloud. Out of the box, however, it is remarkably limited. The software lacks basic conveniences like conversation history, audible responses, or reliable smart home integration. It arrives as a hollow shell, requiring users to bridge the gap between hardware and utility through manual programming.

Using Claude Code, I transformed the device into something far more capable than the stock experience. I integrated an ElevenLabs AI voice, enabled a remote camera viewfinder, and coded a persistent Tamagotchi-style minigame where avatars require food, sleep, and hygiene. This process required no prior coding experience, proving that the true value of such hardware lies in its malleability rather than its factory state. While Meta’s proprietary Muse Charm promises a polished experience, the ability to inject custom 'vibe code' into a $50 devkit highlights a critical question: will the official product allow for this level of personal control, or will it remain a locked-down accessory compared to the versatility of open-source alternatives?