Viz MCP Server 📊

A Model Context Protocol (MCP) server for data visualization — generate bar charts, line charts, pie charts, scatter plots, and histograms. Returns SVG (inline) or PNG (file).

Built for AI agents. Works with Hermes Agent, Claude Code, Cursor, and any MCP-compatible client.

✨ Features

Tool	Description
create_bar_chart	Bar chart (vertical/horizontal) with value labels
create_line_chart	Line chart with area fill, markers, trend
create_pie_chart	Pie/donut chart with percentage labels
create_scatter_plot	Scatter plot with optional regression line
create_histogram	Histogram with mean line, optional cumulative

All tools support:

• 4 color palettes: default, vibrant, pastel, monochrome
• SVG output (inline for MCP response) or PNG output (saved to file)
• Custom titles, axis labels
• Clean matplotlib styling (no chartjunk)

🚀 Quick Start

# Install from GitHub
pip install git+https://github.com/ceeyang-ai/viz-mcp-server.git

# Run as MCP server
viz-mcp-server

🔌 Usage with Hermes Agent

Add to ~/.hermes/config.yaml:

mcp_servers:
  viz:
    command: "viz-mcp-server"

Restart Hermes → tools available as mcp_viz_create_bar_chart, etc.

📖 Examples

Bar Chart

# Via MCP tool call
result = create_bar_chart(
    values=[10, 25, 15, 30, 20],
    labels=["Q1", "Q2", "Q3", "Q4", "Q5"],
    title="Quarterly Revenue",
    ylabel="Revenue ($K)",
    palette="vibrant"
)

Scatter with Trend Line

result = create_scatter_plot(
    x_values=[1, 2, 3, 4, 5, 6, 7, 8],
    y_values=[2, 3, 5, 7, 11, 13, 17, 19],
    title="Growth Analysis",
    regression_line=True
)

Histogram

result = create_histogram(
    values=[12, 15, 13, 20, 19, 18, 14, 16, 22, 25, 21, 17],
    bins=8,
    title="Score Distribution",
    xlabel="Score"
)

🛠 Requirements

• Python 3.10+
• matplotlib ≥ 3.7
• numpy ≥ 1.24
• mcp ≥ 1.0

👨‍💻 Development

git clone https://github.com/ceeyang-ai/viz-mcp-server.git
cd viz-mcp-server
pip install -e .
viz-mcp-server  # Start MCP server

📄 License

MIT