leo-claude-mktplace/plugins/viz-platform

viz-platform Plugin

Visualization tools with Dash Mantine Components validation, Plotly charts, and theming for Claude Code.

Features

• DMC Validation: Prevent prop hallucination with version-locked component registry
• Chart Creation: Plotly charts with automatic theme token application
• Layout Builder: Dashboard layouts with filters, grids, and responsive design
• Theme System: Create, extend, and export design tokens

Installation

This plugin is part of the leo-claude-mktplace. Install via:

# From marketplace
claude plugins install leo-claude-mktplace/viz-platform

# Setup MCP server venv
cd ~/.claude/plugins/marketplaces/leo-claude-mktplace/mcp-servers/viz-platform
python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

Configuration

System-Level (Optional)

Create ~/.config/claude/viz-platform.env for default theme preferences:

VIZ_PLATFORM_COLOR_SCHEME=light
VIZ_PLATFORM_PRIMARY_COLOR=blue

Project-Level (Optional)

Add to project .env for project-specific settings:

VIZ_PLATFORM_THEME=my-custom-theme
DMC_VERSION=0.14.7

Commands

Command	Description
/initial-setup	Interactive setup wizard for DMC and theme preferences
/component {name}	Inspect component props and validation
/chart {type}	Create a Plotly chart
/dashboard {template}	Create a dashboard layout
/theme {name}	Apply an existing theme
/theme-new {name}	Create a new custom theme
/theme-css {name}	Export theme as CSS

Agents

Agent	Description
theme-setup	Design-focused theme creation specialist
layout-builder	Dashboard layout and filter specialist
component-check	Strict component validation specialist

Tool Categories

DMC Validation (3 tools)

Prevent invalid component props before runtime.

Tool	Description
list_components	List available components by category
get_component_props	Get detailed prop specifications
validate_component	Validate a component configuration

Charts (2 tools)

Create Plotly charts with theme integration.

Tool	Description
chart_create	Create a chart (line, bar, scatter, pie, etc.)
chart_configure_interaction	Configure chart interactivity

Layouts (5 tools)

Build dashboard structures with filters and grids.

Tool	Description
layout_create	Create a layout structure
layout_add_filter	Add filter components
layout_set_grid	Configure responsive grid
layout_add_section	Add content sections
layout_get	Retrieve layout details

Themes (6 tools)

Manage design tokens and styling.

Tool	Description
theme_create	Create a new theme
theme_extend	Extend an existing theme
theme_validate	Validate theme configuration
theme_export_css	Export as CSS custom properties
theme_list	List available themes
theme_activate	Set the active theme

Pages (5 tools)

Create full Dash app configurations.

Tool	Description
page_create	Create a page structure
page_add_navbar	Add navigation bar
page_set_auth	Configure authentication
page_list	List pages
page_get_app_config	Get full app configuration

Component Validation

The key differentiator of viz-platform is the component registry system:

# Before writing component code
get_component_props("Button")
# Returns: all valid props with types, enums, defaults

# After writing code
validate_component("Button", {"variant": "filled", "color": "blue"})
# Returns: {valid: true} or {valid: false, errors: [...]}

This prevents common DMC mistakes:

• Prop typos (colour vs color)
• Invalid enum values (size="large" vs size="lg")
• Wrong case (fullwidth vs fullWidth)

Example Workflow

/component Button
# → Shows all Button props with types and defaults

/theme-new corporate
# → Creates theme with brand colors

/chart bar
# → Creates bar chart with theme colors

/dashboard sidebar
# → Creates sidebar layout with filters

/theme-css corporate
# → Exports theme as CSS for external use

Cross-Plugin Integration

viz-platform works seamlessly with data-platform:

1. Load data with data-platform: /ingest sales.csv
2. Create chart with viz-platform: /chart line using the data_ref
3. Build layout with viz-platform: /dashboard with filters
4. Export complete dashboard structure

Chart Types

Type	Best For
line	Time series, trends
bar	Comparisons, categories
scatter	Correlations, distributions
pie	Part-to-whole
area	Cumulative trends
histogram	Frequency distributions
box	Statistical distributions
heatmap	Matrix correlations
sunburst	Hierarchical data
treemap	Hierarchical proportions

Layout Templates

Template	Best For
basic	Simple dashboards, reports
sidebar	Navigation-heavy apps
tabs	Multi-page dashboards
split	Comparisons, master-detail

Requirements

• Python 3.10+
• dash-mantine-components >= 0.14.0
• plotly >= 5.18.0
• dash >= 2.14.0