What is this?
VoiceShell Muse Bridge is an experimental research project that lets an ESP32 board pair with Meta's Muse Gadget SDK, then hands the actual text conversation off to a separate Linux server instead of routing every message through the board.
VoiceShell Muse Bridge is an experimental research project that lets an ESP32 board pair with Meta's Muse Gadget SDK, then hands the actual text conversation off to a separate Linux server instead of routing every message through the board.Is this a finished app or just research code?It is developer research code documenting reproducible experiments and dead ends, not a packaged product or one-command installer.
It is developer research code documenting reproducible experiments and dead ends, not a packaged product or one-command installer.How do I try this out?You pick an entry point from a table (ESP32 setup, Serial bridge, Linux migration, output-queue client, or compatibility history) and run the credential-free tests from the repository root.
You pick an entry point from a table (ESP32 setup, Serial bridge, Linux migration, output-queue client, or compatibility history) and run the credential-free tests from the repository root.What do I need before running the Linux setup?Python 3.9+, an SDK token, the official Muse app, and user-approved device pairing on a tested ESP32-S3 board.
Python 3.9+, an SDK token, the official Muse app, and user-approved device pairing on a tested ESP32-S3 board.Does BLE pairing just work out of the box?No. The authors hit real failures on both Mac and Windows before a physical ESP32-S3 finally showed up in the Muse app, and Wi-Fi/cloud connectivity issues can look identical to authentication failures.
No. The authors hit real failures on both Mac and Windows before a physical ESP32-S3 finally showed up in the Muse app, and Wi-Fi/cloud connectivity issues can look identical to authentication failures.Do I need a Linux machine and an ESP32 board to use this?Yes — the project pairs through a physical ESP32-S3 and then runs the real conversation on a Linux server; Mac and Windows are only documented as failed BLE attempts, not supported paths.
Yes — the project pairs through a physical ESP32-S3 and then runs the real conversation on a Linux server; Mac and Windows are only documented as failed BLE attempts, not supported paths.What happens to my credentials and tokens?Tokens, pairing state, and credential-bearing firmware binaries are kept out of Git, and the Linux experiment only exposes a minimal text-reply tool, not the SDK's default shell/file access or OTA.
Tokens, pairing state, and credential-bearing firmware binaries are kept out of Git, and the Linux experiment only exposes a minimal text-reply tool, not the SDK's default shell/file access or OTA.What is Muse, and why does this project care about it?Muse Gadgets is Meta's wearable device platform with an official SDK and companion app; the project tests whether the team's own hardware can pair with it, noting that community devices are not verified by Meta and get broad agent access.
Muse Gadgets is Meta's wearable device platform with an official SDK and companion app; the project tests whether the team's own hardware can pair with it, noting that community devices are not verified by Meta and get broad agent access.Why build a bridge instead of just using the ESP32 directly?The team's own VoiceShell headset isn't an ESP32, so they wanted to confirm an ESP32 could handle initial pairing while a Linux server runs the actual conversation, letting future adapters be built around the server rather than a development board.
The team's own VoiceShell headset isn't an ESP32, so they wanted to confirm an ESP32 could handle initial pairing while a Linux server runs the actual conversation, letting future adapters be built around the server rather than a development board.Is this ready for everyday, always-on use?No. Sessions were closed manually after short tests (19.4s and 6.6s), and persistent service, token renewal, multi-user isolation, and full headset audio (ASR to Muse to TTS) are all still on the roadmap.
No. Sessions were closed manually after short tests (19.4s and 6.6s), and persistent service, token renewal, multi-user isolation, and full headset audio (ASR to Muse to TTS) are all still on the roadmap.What is VoiceShell itself?VoiceShell (声壳) is wearable voice-input hardware and companion software for quieter, less conspicuous conversations with personal agents; this Muse bridge is one experiment toward that larger goal.
VoiceShell (声壳) is wearable voice-input hardware and companion software for quieter, less conspicuous conversations with personal agents; this Muse bridge is one experiment toward that larger goal.What are similar projects?xiaozhi-esp32, mimiclaw, hermes-gadget-sdk, RuView, EchoMuse, techo5, arduino-esp32, muse-gadget-sdk
xiaozhi-esp32, mimiclaw, hermes-gadget-sdk, RuView, EchoMuse, techo5, arduino-esp32, muse-gadget-sdkWhich topics is it in?muse, ai-agent-tools, bluetooth, agent-tools, esp32, voice-assistant, embedded-systems, firmware, iot, wechat, embedded, hardware
muse, ai-agent-tools, bluetooth, agent-tools, esp32, voice-assistant, embedded-systems, firmware, iot, wechat, embedded, hardware