Halo PixelBar MCP
A local MCP server for controlling an EDIFIER Halo PixelBar's ambient light, pixel screen, and hardware volume on Windows.
README
Halo PixelBar MCP
A local MCP server for controlling an EDIFIER Halo PixelBar from an AI on Windows. It exposes a small allowlist for the ambient light, pixel screen, and the speaker's independent hardware volume.
This is an independent community project. It is not affiliated with, endorsed by, or supported by EDIFIER.
See PROTOCOL_NOTES.md for the observed packet format, field mappings, and live hardware tests.
Prerequisites
- Windows 10 or Windows 11.
- An EDIFIER Halo PixelBar (花再) exposing USB VID
0x2D99, PID0xA106, HID interface4, usage page0xFF14, and usage1. - A USB connection that carries data, not a power-only cable. A data-capable adapter or dock is fine when Windows can enumerate the HID interface.
- Python 3.11 or newer.
- A local MCP host that can launch a stdio server, such as Codex or Claude Desktop.
EDIFIER TempoHub was installed while the protocol was validated, but this MCP does not call TempoHub, read an EDIFIER account, or require account credentials. The device filter is intentionally strict; other EDIFIER speakers and PixelBar revisions are not assumed to be compatible.
Install
git clone https://github.com/Seraph310/halo-pixelbar-mcp.git
cd halo-pixelbar-mcp
py -3.11 -m venv .venv
.\.venv\Scripts\python.exe -m pip install -e ".[dev]"
Codex configuration
Add the following to %USERPROFILE%\.codex\config.toml, replacing both paths:
[mcp_servers.halo_pixelbar]
command = 'C:\path\to\halo-pixelbar-mcp\.venv\Scripts\python.exe'
args = ['C:\path\to\halo-pixelbar-mcp\server.py']
startup_timeout_sec = 30
Restart the MCP host after changing its configuration.
Tools
| Tool | Purpose |
|---|---|
get_halo_status |
Detect the validated HID control interface |
get_halo_volume |
Read the independent 0-16 speaker volume |
set_halo_volume |
Set the hardware volume and verify it by querying again |
show_halo_text |
Display a short UTF-8 message with a safe layout |
set_halo_scene |
Select a built-in clock, game, work, reading, pet, meme, cyber, or waveform scene |
set_halo_screen_color |
Set the pixel-screen theme color |
set_halo_ambient |
Set light effect, RGB color, brightness, and speed |
set_halo_ambient_power |
Switch only the ambient light on or off |
Tested environment
- Windows 11 64-bit, build
26200. - EDIFIER Halo PixelBar (花再), USB VID
0x2D99, PID0xA106. - EDIFIER TempoHub installer version
1.2.6. - Python
3.11.5. - A data-capable USB adapter; a direct motherboard connection is not required when Windows can still enumerate the HID interface.
Safety boundary
- Targets only the exact validated VID, PID, interface, usage page, usage, and a matching Halo/PixelBar product name.
- Exposes no arbitrary HID packet, firmware, factory-reset, account, file, or shell-command tool.
- Validates colors, scenes, effects, speed, brightness, layouts, text length, and hardware volume before writing.
- Does not require or read EDIFIER TempoHub credentials.
- Omits the physical device serial number from public MCP status results.
Optional music companion
Pair this server with CloudMusic Desktop MCP to let an MCP host coordinate AI-selected music with light, screen, and hardware-volume changes. The two servers are independent and can be installed separately.
Protocol attribution
The packet format was established through local device observation, the installed official controller, and live readback tests. It was also compared with the MIT-licensed XFEstudio/HaloPixelToolBox implementation and the official Halo PixelBar user manual. No extracted TempoHub binaries, decompiled files, credentials, firmware, or research snapshots are included in this repository.
Test
.\.venv\Scripts\python.exe -m pytest -q
Credits
Built together by Celia and OpenAI Codex.
Celia led the product direction and live hardware validation. Codex assisted with reverse engineering, implementation, and testing.
License
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。