Vizro MCP

Name: Vizro MCP
Author: mckinsey

Verified

Tools and templates to create validated and maintainable data charts and dashboards.

by mckinsey Official 3,634 Apache-2.0 Python Intermediate stdio

Install

Claude Code

claude mcp add vizro -- uvx vizro-mcp

Cursor

{
  "mcpServers": {
    "vizro": {
      "command": "uvx",
      "args": [
        "vizro-mcp"
      ]
    }
  }
}

VS Code

{
  "mcp": {
    "servers": {
      "vizro": {
        "command": "uvx",
        "args": [
          "vizro-mcp"
        ]
      }
    }
  }
}

Windsurf

{
  "mcpServers": {
    "vizro": {
      "command": "uvx",
      "args": [
        "vizro-mcp"
      ]
    }
  }
}

Claude Desktop

{
  "mcpServers": {
    "vizro": {
      "command": "uvx",
      "args": [
        "vizro-mcp"
      ]
    }
  }
}

Safety Report

100% Verified

5 passed

Security

Authentication

details

No MCP server authentication required. Authentication is handled by the LLM API provider (e.g., Anthropic, OpenAI) configured separately.

Info

CORS Policy

details

Local stdio transport. CORS not applicable.

N/A

Rate Limiting

details

Local processing only. No external API calls from the MCP server itself.

N/A

Known CVEs

details

No known CVEs. Enterprise-backed project with McKinsey engineering oversight.

Pass

Code Quality

Dependency Audit

details

Part of the Vizro monorepo maintained by McKinsey & Company. Python package with standard dependency management.

Pass

Code Source

details

Open source on GitHub (mckinsey/vizro). Professionally maintained by McKinsey's engineering team.

Pass

License

details

Apache License 2.0 (Vizro project standard). Explicitly confirmed in monorepo.

Pass

Community

Maintenance Status

details

Actively maintained by McKinsey & Company. Part of a well-resourced open source project with regular releases.

Pass

Learn about our security methodology →

Compatibility

Claude CodeSupported via uv or Docker. Recommended to use claude-sonnet-4 or better model for best results.

Not Tested

CursorDocumented support via .cursor/mcp.json with uv command configuration.

Not Tested

VS CodeCompatible with VS Code MCP extensions using standard stdio transport.

Not Tested

WindsurfStandard MCP JSON configuration supported.

Not Tested

Claude DesktopSupported via claude_desktop_config.json with uv or Docker command.

Not Tested

Tools Provided (6)

get_vizro_chart_or_dashboard_plan

Get step-by-step instructions for creating a Vizro chart or dashboard. Should be called first when asked to create any Vizro visualization.

get_model_json_schema

Get the JSON schema definition for a specified Vizro model component.

get_sample_data_info

Get information about built-in sample datasets (iris, tips, stocks, gapminder) when no user data is provided.

load_and_analyze_data

Load and analyze a local or remote data file. Must be called with an absolute file path or URL.

validate_dashboard_config

Validate a complete Vizro dashboard configuration. Should always be run before finalizing a dashboard.

validate_chart_code

Validate chart code and optionally open a PyCafe preview link in the browser.

Frequently Asked Questions

What is Vizro MCP?

Tools and templates to create validated and maintainable data charts and dashboards.

What tools does Vizro MCP provide?

Vizro MCP provides 6 tools including get_vizro_chart_or_dashboard_plan, get_model_json_schema, get_sample_data_info, load_and_analyze_data. Each tool can be called by AI agents to perform specific operations.

Is Vizro MCP safe to use?

Yes, Vizro MCP has passed our 8-point security audit and is rated Verified. All checks including authentication, dependency audit, and known CVE scan passed.

What are alternatives to Vizro MCP?

Similar MCP servers include MCP Server Chart, Markdownify MCP, Jupyter MCP Server. Each serves a similar purpose but may differ in features, language, and compatibility.