Vizdown-MCP
Converts Markdown files into professional diagrams such as flowcharts, mind maps, and architecture diagrams using Mermaid.js and custom SVG renderers. It enables users to list, render, and export visualizations in multiple formats including SVG, PNG, and PDF.
README
Clarity-beta
Clarity-beta is an MCP server that converts Markdown files into beautiful diagrams — flowcharts, mind maps, architecture diagrams, ER diagrams, Gantt charts, sequence diagrams, and more. Built with Apple HIG design language.
Installation
git clone https://github.com/rutika196/clarity-beta.git
cd clarity-beta
python3 -m venv .venv
source .venv/bin/activate # macOS / Linux
# .venv\Scripts\activate # Windows
pip install -e .
playwright install chromium
IDE Setup
Cursor
Create or edit .cursor/mcp.json in your project root:
{
"mcpServers": {
"clarity-beta": {
"command": "/absolute/path/to/clarity-beta/.venv/bin/python",
"args": ["-m", "src.server"],
"cwd": "/absolute/path/to/clarity-beta"
}
}
}
Replace
/absolute/path/to/clarity-betawith your actual project path.
VS Code
Create .vscode/mcp.json in your project root:
{
"servers": {
"clarity-beta": {
"command": "/absolute/path/to/clarity-beta/.venv/bin/python",
"args": ["-m", "src.server"],
"cwd": "/absolute/path/to/clarity-beta"
}
}
}
VS Code reads MCP config from
.vscode/mcp.json— no need to touchsettings.json.
IntelliJ IDEA / WebStorm / PyCharm (JetBrains)
JetBrains IDEs (2025.1+) support MCP via the AI Assistant plugin.
- Go to Settings → Tools → AI Assistant → MCP Servers.
- Click + Add and fill in:
| Field | Value |
|---|---|
| Name | clarity-beta |
| Command | /absolute/path/to/clarity-beta/.venv/bin/python |
| Arguments | -m src.server |
| Working Directory | /absolute/path/to/clarity-beta |
Or edit the MCP config file directly at ~/.config/jetbrains/mcp.json:
{
"mcpServers": {
"clarity-beta": {
"command": "/absolute/path/to/clarity-beta/.venv/bin/python",
"args": ["-m", "src.server"],
"cwd": "/absolute/path/to/clarity-beta"
}
}
}
Windows users: replace
.venv/bin/pythonwith.venv\\Scripts\\python.exeand use\\in paths.
Claude Desktop
Edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):
{
"mcpServers": {
"clarity-beta": {
"command": "/absolute/path/to/clarity-beta/.venv/bin/python",
"args": ["-m", "src.server"],
"cwd": "/absolute/path/to/clarity-beta"
}
}
}
Usage
As a standalone server (stdio transport)
source .venv/bin/activate
clarity-beta
# or
python -m src.server
MCP Tools
render_diagram
Render the first diagram found in a Markdown file or raw text.
| Parameter | Type | Default | Description |
|---|---|---|---|
file_path |
string | — | Path to a .md file |
raw_markdown |
string | — | Raw markdown with a diagram block |
output_format |
string | "svg" |
"svg", "png", "jpeg", or "pdf" |
theme |
string | "light" |
"light" or "dark" |
look |
string | "default" |
"default" or "handDrawn" |
scale |
int | 2 |
Resolution multiplier for PNG/JPEG |
render_all_diagrams
Render every diagram block in a file. Same parameters plus output_dir to save files.
list_diagrams
List detected diagram blocks with types and line numbers, without rendering.
Supported Diagram Types
Via Mermaid.js (Playwright)
flowchart, sequence, class, ER, state, gantt, gitGraph, pie, timeline, quadrant, sankey, xychart, block-beta, architecture-beta, kanban, journey, C4
Custom SVG Renderers
- Mind maps — balanced horizontal tree with organic Bézier connectors
- Architecture / service diagrams — Miro-style auto-layout with icons, step badges, groups
Examples
See the examples/ folder — 13 ready-to-render Markdown files:
| File | Diagram Type |
|---|---|
auth_flow.md |
Flowchart (OAuth 2.0 + MFA) |
api_sequence.md |
Sequence diagram |
order_states.md |
State diagram |
database_schema.md |
ER diagram |
class_diagram.md |
Class diagram |
project_roadmap.md |
Gantt chart |
tech_stack_pie.md |
Pie chart |
git_workflow.md |
Git graph |
company_timeline.md |
Timeline |
user_journey.md |
User journey |
system_overview.md |
Mind map (custom SVG) |
microservices.md |
Architecture (17 services + groups) |
ci_cd_pipeline.md |
Architecture (CI/CD pipeline) |
Running Tests
python tests/test_all.py
Dependencies
mcp[cli]— MCP Python SDKplaywright— Headless Chromium for Mermaid rendering and PNG/JPEG exportcairosvg— SVG to PDF conversionPillow— Image processing (optional, for JPEG)
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。