symbols-mcp-server
An MCP server that exposes the Symbols/DOMQL v3 AI assistant capabilities to any MCP-compatible platform. Enables generating components, pages, projects, and more from natural language, as well as searching documentation and reviewing code.
README
Symbols MCP Server
An MCP (Model Context Protocol) server that exposes the Symbols/DOMQL v3 AI assistant capabilities to any MCP-compatible platform — Cursor, Claude Code, Windsurf, and more.
Features
Tools
| Tool | Description |
|---|---|
generate_component |
Generate a Symbols/DOMQL v3 component from natural language |
generate_page |
Generate a full page with routing support |
generate_project |
Scaffold a complete multi-file Symbols project |
convert_to_symbols |
Convert React/Angular/Vue/HTML to Symbols/DOMQL v3 |
search_symbols_docs |
Search Symbols documentation (vector or local) |
explain_symbols_concept |
Explain any Symbols concept with examples |
review_symbols_code |
Review code for v3 compliance and best practices |
create_design_system |
Generate design system files (colors, spacing, themes, icons) |
Resources
| Resource URI | Description |
|---|---|
symbols://skills/domql-v3-reference |
Complete DOMQL v3 syntax reference |
symbols://skills/project-structure |
Project folder structure conventions |
symbols://skills/design-direction |
Modern UI/UX design direction |
symbols://skills/migration-guide |
React/Angular/Vue → Symbols migration |
symbols://skills/v2-to-v3-migration |
DOMQL v2 → v3 changes |
symbols://skills/quickstart |
CLI setup and quickstart |
symbols://reference/spacing-tokens |
Spacing token reference table |
symbols://reference/atom-components |
Built-in primitive components |
symbols://reference/event-handlers |
Event handler reference |
Prompts
| Prompt | Description |
|---|---|
symbols_component_prompt |
Template for component generation |
symbols_migration_prompt |
Template for framework migration |
symbols_project_prompt |
Template for project scaffolding |
symbols_review_prompt |
Template for code review |
Installation
Quick Start (No API Keys Required!)
- Clone the repository:
git clone https://github.com/baronsilver/symbols-mcp-server.git
cd symbols-mcp-server
- Install dependencies:
pip install uv
uv sync
npm i @symbo.ls/cli -g
smbls create your-project
IMPORTANT: For now, delete the docs folder inside your project folder after creation
- Configure:
cp .env.example .env
# Edit .env and add both:
# - SYMBOLS_MCP_URL (contact maintainer for public server URL)
# - LLM_MODEL=google/gemini-3-flash-preview
Platform Integration
Claude Code
claude mcp add symbols-mcp -- uv run --directory C:\repos\symbols-mcp-server symbols-mcp
Or manually edit ~/UserName/.claude.json:
{
"mcpServers": {
"symbols-mcp": {
"command": "python",
"args": [
"-m",
"uv",
"run",
"--directory",
"C:\\repos\\symbols-mcp-server",
"symbols-mcp"
]
}
}
}
Cursor
Add to your Cursor MCP settings (.cursor/mcp.json in your project or global settings):
{
"mcpServers": {
"symbols-mcp": {
"command": "python",
"args": [
"-m",
"uv",
"run",
"--directory",
"C:\\repos\\symbols-mcp-server",
"symbols-mcp"
]
}
}
}
Windsurf
Add to ~/.codeium/windsurf/mcp_config.json:
{
"mcpServers": {
"symbols-mcp": {
"command": "python",
"args": [
"-m",
"uv",
"run",
"--directory",
"C:\\repos\\symbols-mcp-server",
"symbols-mcp"
]
}
}
}
Development
Restart your platforms after updating the MCP settings. Once loaded, it should automatically run the MCP. Ask your AI to do the quick test with the MCP to see if it works properly.
Environment Variables
| Variable | Required | Default | Description |
|---|---|---|---|
OPENROUTER_API_KEY |
Yes | — | API key for OpenRouter (powers AI generation) |
SUPABASE_URL |
No | — | Supabase project URL for vector search |
SUPABASE_KEY |
No | — | Supabase service role key |
LLM_MODEL |
No | openai/gpt-4.1-mini |
AI model to use via OpenRouter |
SYMBOLS_SKILLS_DIR |
No | ./symbols_mcp/skills |
Path to skills markdown files |
Architecture
symbols-mcp-server/
├── pyproject.toml # Project config and dependencies
├── .env.example # Environment variable template
├── README.md # This file
└── symbols_mcp/
├── __init__.py
├── server.py # MCP server with tools, resources, prompts
└── skills/ # Bundled Symbols knowledge base
├── CLAUDE.md # DOMQL v3 complete reference
├── SYMBOLS_LOCAL_INSTRUCTIONS.md # Project structure rules
├── DESIGN_DIRECTION.md # UI/UX design direction
├── MIGRATE_TO_SYMBOLS.md # Framework migration guide
├── DOMQL_v2-v3_MIGRATION.md # v2→v3 changes
└── QUICKSTART.md # CLI quickstart
The server reads skills files at startup and uses them as context for all AI-powered tools.
When Supabase is configured, the search_symbols_docs tool also queries the vector database
for additional documentation matches.
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。