dct-chart - SKILL.md Agent Skill

name: dct-chart description: Use this skill when the user wants to visualize data distributions, create ASCII histograms, generate simple charts from CSV/JSON data, plot column values, or see value frequencies in terminal-friendly format. Triggers include "chart this data", "visualize distribution", "histogram of values", "plot the data", "ascii chart", "terminal visualization", or when needing quick visual analysis without external plotting tools.

DCT Chart - ASCII Data Visualization

Generate simple ASCII histograms from data file columns.

When to Use

Use this skill when you need to:

Quickly visualize value distributions
Create terminal-friendly charts
Analyze categorical data frequencies
Get a quick histogram without external tools
Share visualizations in text format

Installation

which dct || go build -o dct && chmod +x ./dct

Usage

dct chart [file] [column_index] [flags]

Arguments

file: Data file path (optional)
column_index: 0-based column index to chart (default: 0)

Flags

-w, --width <chars>: Width of chart in characters (default: auto)

Examples

Basic Usage

Chart first column:

dct chart data.csv

Chart specific column (0-based index):

dct chart data.csv 2

Custom Width

Wider chart:

dct chart sales.csv 1 -w 80

Compact chart:

dct chart data.csv 0 -w 40

Different File Types

Chart from Parquet:

dct chart data.parquet 3

Chart from JSON:

dct chart data.json 0

Output Format

Displays an ASCII histogram with:

Column values on the left (Y-axis labels)
Bar representation using box-drawing characters
Count/frequency on the right
Box border around the chart

Example output:

            ┌─ histogram of 'sales.csv' ──────────────────────────┐
    ProductA ┤ ╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍ 156 │
    ProductB ┤ ╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍ 98 │
    ProductC ┤ ╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍ 73 │
    ProductD ┤ ╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍ 45 │
    ProductE ┤ ╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍╍ 22 │
             └───────────────────────────────────────────────────┘

How It Works

Reads the specified column from the data file
Groups values and counts frequencies
Sorts by frequency (descending)
Renders ASCII bar chart scaled to fit terminal

Best Practices

Column indexing: Remember column index is 0-based (first column = 0)
Use -w flag: Adjust width to fit your terminal
Categorical data: Works best with categorical or binned data
High cardinality: Top values are shown; long tail may be truncated
Preview first: Use dct-peek to see column names before charting

Use Cases

Sales Analysis

# Chart sales by product
dct chart sales.csv 0 -w 60

# Chart revenue distribution
dct chart orders.csv 2 -w 80

Data Exploration

# Quick distribution check
dct peek data.csv -n 5
dct chart data.csv 1 -w 50

Log Analysis

# Chart status codes from logs
dct chart logs.csv 3 -w 40

Limitations

Shows top N values (high-cardinality columns truncate)
ASCII art has limited resolution
Best for categorical or discrete numeric data
Continuous numeric data should be pre-binned

Column Selection Workflow

# 1. Preview file to see columns
dct peek data.csv

# 2. Note the column index (0-based)
#    Columns: id | name | category | amount
#    Index:    0    1        2         3

# 3. Chart desired column
dct chart data.csv 2 -w 60

Integration Examples

In Shell Scripts

#!/bin/bash
# Generate charts for all columns in a file
file="data.csv"
num_cols=$(head -1 "$file" | tr ',' '\n' | wc -l)
for i in $(seq 0 $((num_cols - 1))); do
    echo "Column $i:"
    dct chart "$file" $i -w 50
    echo
done

With Other Tools

# Filter data then chart
grep "ERROR" logs.csv | dct chart - 1 -w 40

# Save chart to file
dct chart data.csv 0 -w 80 > chart.txt

Related Skills

dct-peek: Preview column names before charting
dct-profile: Get detailed statistics about the column
dct-infer: Understand column types

Terminal Tips

Charts look best in terminals with UTF-8 support
Use wider terminals (-w 100+) for detailed views
Pipe to less -S for wide charts: dct chart data.csv 0 -w 120 | less -S