/data-viz

allowed-tools: Read, Write, Edit, MultiEdit, Task, Bash(fd:*), Bash(rg:*), Bash(jq:*), Bash(gdate:*), Bash(wc:*)
name: "Data Viz"
description: "Generate interactive data visualizations and dashboards with intelligent chart selection and real-time updates"
author: "wcygan"
tags: ["analyze","data"]
version: "1.0.0"
created_at: "2025-07-14T00:00:00Z"
updated_at: "2025-07-14T00:00:00Z"

Context

• Session ID: !`gdate +%s%N`
• Current directory: !`pwd`
• Data files: !`fd "\.(csv|json|xlsx|parquet|tsv|xml)$" --max-depth 3 | head -10 || echo "No data files found"`
• Database configs: !`fd "(knex|prisma|typeorm|sequelize|deno\.json|package\.json)" --max-depth 2 | head -5 || echo "No database configs found"`
• API endpoints: !`rg "/(api|data|metrics|analytics)/" --type js --type ts --type go --type rust | head -10 || echo "No API endpoints found"`
• Technology stack: !`fd "(deno\.json|package\.json|Cargo\.toml|go\.mod)" --max-depth 2 | head -5 || echo "No framework files detected"`
• Existing dashboards: !`rg "(dashboard|chart|visualization|d3|chartjs)" --type js --type ts | head -5 || echo "No existing dashboards found"`
• Analytics tools: !`fd "(grafana|kibana|metabase|analytics)" --type d | head -5 || echo "No analytics tools found"`
• Data volume estimate: !`fd "\.(csv|json|xlsx)$" --max-depth 3 -x wc -l {} \; 2>/dev/null | head -5 || echo "No data volume info"`

Your Task

Generate interactive data visualizations and dashboards for: **$ARGUMENTS**

STEP 1: Session Initialization and State Management

• Initialize session state file: /tmp/data-viz-state-$SESSION_ID.json
• Create temporary workspace: /tmp/data-viz-workspace-$SESSION_ID/
• Set up analysis tracking and progress monitoring

// /tmp/data-viz-state-$SESSION_ID.json
{
  "sessionId": "$SESSION_ID",
  "timestamp": "ISO_8601_TIMESTAMP",
  "target": "$ARGUMENTS",
  "phase": "initialization",
  "discovery": {
    "dataFiles": [],
    "apiEndpoints": [],
    "databaseSchemas": [],
    "existingDashboards": []
  },
  "analysis": {
    "dataTypes": {},
    "chartRecommendations": [],
    "frameworkChoice": null,
    "complexity": "simple|moderate|complex"
  },
  "generation": {
    "components": [],
    "assets": [],
    "configurations": []
  },
  "checkpoints": {
    "discovery_complete": false,
    "analysis_complete": false,
    "generation_complete": false,
    "validation_complete": false,
    "deployment_ready": false
  }
}

STEP 2: Data Discovery and Schema Analysis

IF $ARGUMENTS provided:

• Analyze specified data source (file, URL, or description)
• Skip comprehensive discovery and focus on target analysis

ELSE:

• Execute comprehensive data source discovery

Think deeply about optimal data discovery strategies and visualization opportunities for this project.

Use parallel sub-agents for comprehensive data discovery:

• **Agent 1**: File Data Discovery and Schema Analysis
• Analyze CSV, JSON, Excel files for structure and content
• Determine data types, ranges, and relationships
• Identify temporal patterns and categorical distributions
• Extract sample data for chart recommendations

• **Agent 2**: API Data Discovery and Integration Analysis
• Discover REST/GraphQL endpoints returning data
• Test endpoint schemas and response formats
• Analyze real-time data capabilities and update frequencies
• Document authentication and access requirements

• **Agent 3**: Database Schema Analysis
• Detect database configurations and connection strings
• Analyze table schemas and relationship mappings
• Identify time-series tables and aggregation opportunities
• Evaluate query performance and optimization needs

• **Agent 4**: Existing Analytics Infrastructure Assessment
• Find existing dashboards, charts, and visualization tools
• Analyze current visualization libraries and frameworks
• Identify integration opportunities and migration paths
• Assess team preferences and technical constraints

TRY:

• Execute parallel data discovery across all available sources
• Consolidate findings into comprehensive data inventory
• Update state: discovery_complete = true, phase = "analysis"

CATCH (no_data_sources_found):

• Generate sample datasets for demonstration purposes
• Create mock API endpoints for prototype development
• Document ideal data source requirements
• Update state with fallback data generation plan

STEP 3: Intelligent Chart Recommendation and Framework Selection

Think harder about visualization design principles and optimal chart selection strategies for the discovered data patterns.

PROCEDURE analyze_data_characteristics():

• FOR EACH discovered data source:
• Classify data types: numerical, categorical, temporal, geographical, hierarchical
• Calculate data volumes and update frequencies
• Identify key relationships and correlation opportunities
• Determine aggregation levels and drill-down possibilities

PROCEDURE recommend_visualizations():

• FOR EACH data characteristic pattern:

**Temporal Data Patterns**:

• IF time_series_data: RECOMMEND line charts, area charts, time heatmaps
• IF seasonal_patterns: RECOMMEND calendar heatmaps, cycle plots
• IF real_time_streams: RECOMMEND live updating dashboards

**Categorical Data Patterns**:

• IF few_categories (<=10): RECOMMEND bar charts, pie charts, treemaps
• IF many_categories (>10): RECOMMEND horizontal bars, word clouds, sunburst
• IF hierarchical_categories: RECOMMEND treemaps, sankey diagrams

**Numerical Data Patterns**:

• IF distributions: RECOMMEND histograms, box plots, violin plots
• IF correlations: RECOMMEND scatter plots, correlation matrices
• IF multi_dimensional: RECOMMEND parallel coordinates, radar charts

**Geographical Data Patterns**:

• IF coordinates: RECOMMEND scatter maps, choropleth maps
• IF regions: RECOMMEND filled maps, symbol maps
• IF movement: RECOMMEND flow maps, animated paths

PROCEDURE select_optimal_framework():

• Detect project technology stack f