omarchy-mcp
Enables semantic search over Omarchy, Arch Linux, and Hyprland documentation, allowing users to query configuration and usage information via natural language in Cursor, OpenCode, or Claude Code.
README
Omarchy MCP Server
A Model Context Protocol (MCP) server providing semantic search over Omarchy, Arch Linux, and Hyprland documentation.
Version Information
- Omarchy: v3.8.2 (pinned)
- Omarchy Releases: All versions up to v3.8.2
- Arch Wiki: Latest (updated via script)
- Hyprland Wiki: Latest (updated via script)
Note: This server contains Omarchy v3.8.2 documentation. Some features may differ if you are on a different version.
Prerequisites
- Arch-based Linux system (uses pacman for arch-wiki-docs)
- Docker and Docker Compose
- Git
- 10 GB free disk space
- Internet connection for initial setup
Quick Start
1. Clone the Repository
git clone https://github.com/Zeus-Deus/omarchy-mcp.git
cd omarchy-mcp
2. Run Setup
chmod +x scripts/setup.sh
./scripts/setup.sh
This will take approximately 3-4 minutes and will:
- Restore Omarchy v3.8.2 docs from snapshot
- Download latest Arch Wiki and Hyprland documentation
- Build and start Docker containers
- Process and ingest all documentation into vector database
- Create 8,500+ searchable documentation chunks
Upgrading Omarchy Version
To upgrade to a new Omarchy version:
./scripts/upgrade.sh 3.4.2
docker-compose restart mcp-server
That's it - the script handles everything (download docs, process, ingest, create snapshot).
3. Configure Cursor IDE
Create or edit ~/.cursor/mcp.json:
{
"mcpServers": {
"omarchy-kb": {
"command": "docker",
"args": [
"exec",
"-i",
"omarchy-mcp-server",
"python",
"/app/mcp_server/main.py"
]
}
}
}
Important: Restart Cursor completely after adding the configuration.
4. Configure OpenCode
Create or edit ~/.config/opencode/opencode.json:
{
"$schema": "https://opencode.ai/config.json",
"autoupdate": false,
"mcp": {
"omarchy-kb": {
"type": "local",
"command": [
"docker",
"exec",
"-i",
"omarchy-mcp-server",
"python",
"/app/mcp_server/main.py"
]
}
}
}
5. Configure Claude Code
Run the following command in your terminal:
claude mcp add --scope user omarchy-kb -- docker exec -i omarchy-mcp-server python /app/mcp_server/main.py
Restart Claude Code after adding the configuration. Verify the server is connected:
claude mcp list
6. Use in Cursor / OpenCode / Claude Code
- Open your IDE (Cursor, OpenCode, or Claude Code)
- Switch to Agent mode (click "Ask" dropdown in bottom-left, select "Agent")
- In the chat, type queries like:
Use omarchy-kb to search for waybar configuration
Or:
How do I configure Hyprland keybindings in Omarchy?
Or:
Use omarchy-kb to tell me what's new in Omarchy version 3.2.0
Available Tools
The MCP server provides these tools:
- search_documentation - Semantic search across all documentation
- find_config_location - Find configuration file paths for applications
- compare_omarchy_vs_arch - Compare Omarchy vs vanilla Arch/Hyprland implementations
- get_server_info - View server statistics and capabilities
Documentation Sources
The knowledge base includes:
| Source | Priority | Description |
|---|---|---|
| Omarchy | 1 (highest) | Omarchy-specific documentation and customizations |
| Omarchy Releases | 1 (highest) | GitHub release notes with changelogs and new features |
| Hyprland | 2 | Hyprland window manager documentation |
| Arch Wiki | 3 | Base Arch Linux documentation |
Priority System: When conflicts occur, Omarchy documentation takes precedence over Hyprland and Arch.
Updating Documentation
Note: Omarchy documentation and release notes remain at v3.8.2 (pinned version). To update Arch Wiki and Hyprland documentation, re-run the setup script or manually execute the download and processing scripts.
Manual Operations
Stop the Server
docker-compose down
Start the Server
docker-compose up -d
View Logs
docker logs -f omarchy-mcp-server
Rebuild After Code Changes
docker-compose down
docker-compose build
docker-compose up -d
Project Structure
omarchy-mcp/
├── data/
│ ├── snapshots/
│ │ └── omarchy-3.8.2-processed/ # Version snapshot (in Git)
│ ├── raw/ # Downloaded HTML (ignored)
│ └── processed/ # Cleaned JSON (ignored)
├── scripts/
│ ├── setup.sh # Initial setup script
│ ├── 1_download_archwiki.sh # Download Arch Wiki
│ ├── 2_download_hyprland.sh # Download Hyprland wiki
│ ├── 3_download_omarchy.sh # Download Omarchy manual
│ ├── 4_clean_archwiki.py # Clean Arch HTML to JSON
│ ├── 5_clean_hyprland.py # Clean Hyprland MD to JSON
│ ├── 6_clean_omarchy.py # Clean Omarchy HTML to JSON
│ ├── 7_ingest_to_chroma.py # Ingest to vector database
│ ├── 8_download_omarchy_releases.py # Download Omarchy releases (NEW!)
│ └── 9_clean_omarchy_releases.py # Clean releases to JSON (NEW!)
├── mcp_server/
│ └── main.py # MCP server implementation
├── docker-compose.yml # Docker services definition
├── Dockerfile # Container build instructions
├── requirements.txt # Python dependencies
└── README.md
Troubleshooting
Server Not Connecting in Cursor
- Ensure Docker containers are running:
docker ps - Check server logs:
docker logs omarchy-mcp-server - Verify you are in Agent mode in Cursor (not "Ask" mode)
- Restart Cursor completely after updating config
No Results from Queries
- Verify vector database is populated:
docker exec omarchy-mcp-server python -c "
import chromadb
client = chromadb.HttpClient(host='chromadb', port=8000)
collection = client.get_collection('omarchy_docs')
print(f'Documents: {collection.count()}')
"
- Should show approximately 8,500+ documents
Setup Script Fails
- Ensure Docker daemon is running:
systemctl status docker - Check available disk space:
df -h - Verify internet connection
- Try running steps manually from
scripts/setup.sh
Technical Details
- Vector Database: ChromaDB
- Embedding Model: all-MiniLM-L6-v2 (sentence-transformers)
- Chunk Size: 400 words per document chunk
- Search Method: Cosine similarity on embeddings
- Protocol: Model Context Protocol (MCP) via stdio
Contributing
Issues and pull requests welcome at https://github.com/Zeus-Deus/omarchy-mcp
License
GNU General Public License v3.0 (GPL-3.0)
See LICENSE for full license text.
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。
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。
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。
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。
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。
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。
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。
mcp-server-qdrant
这个仓库展示了如何为向量搜索引擎 Qdrant 创建一个 MCP (Managed Control Plane) 服务器的示例。