chart

name	chart
description	Generate bar, line, pie, or scatter charts from CSV files or inline data. Pipe table query output into chart.
metadata	{"nessi":{"command":"chart","enabled":true}}

Chart

Create data visualizations as SVG. Charts render inline in the chat with a download button.

There are 4 chart types: bar, line, pie, scatter. Each reads data from a CSV file or accepts inline values.

Chart types at a glance

Type	Best for	X axis	Y axis
bar	Comparing categories	Category names	One numeric column
line	Trends over time	Time labels	One or more numeric columns
pie	Proportions of a whole	Category names	One numeric column
scatter	Correlation between two numbers	Numeric column	Numeric column

From CSV file (recommended)

Always use this when data comes from table query. Pass the CSV path as the first argument:

chart bar /output/data.csv --x "region" --y "revenue" --title "Revenue by Region"
chart pie /output/data.csv --x "category" --y "amount" --title "Distribution"
chart line /output/data.csv --x "quarter" --y "revenue,cost,profit" --title "Trends"
chart scatter /output/data.csv --x "frequency" --y "volume" --label "customer" --title "Segmentation"

Options for CSV mode

• --x "column" — the label/category column (required)
• --y "column" — the value column (required). For line, use comma-separated columns for multiple series
• --label "column" — (scatter only) show point names next to dots
• --title "text" — chart title
• --output /output/name.svg — custom output path

From inline data

For quick charts without a file. Use --labels and --values:

chart bar --labels "Jan,Feb,Mar" --values "100,200,150" --title "Sales"
chart pie --labels "Chrome,Firefox,Safari" --values "65,20,15" --title "Browsers"
chart line --labels "Q1,Q2,Q3,Q4" --series '{"Revenue":[100,200,150,250],"Cost":[80,120,100,180]}'

Note: scatter does not support inline mode — it always requires a CSV file.

Common pipeline: table query → chart → present

This is the typical workflow for data analysis:

# Step 1: Aggregate data from a spreadsheet
table query /input/sales.xlsx \
  --select "region, sum(revenue) as Revenue, count() as Orders" \
  --group "region" --sort "Revenue desc" \
  --output /output/by-region.csv

# Step 2: Create chart from the aggregated CSV
chart bar /output/by-region.csv --x "region" --y "Revenue" --title "Revenue by Region"

# Step 3: Display inline with download button
present /output/bar-chart.svg

Scatter example with calc():

table query /input/data.xlsx \
  --select "customer, count() as Frequency, sum(amount) as Volume" \
  --group "customer" --limit 20 \
  --output /output/segments.csv

chart scatter /output/segments.csv --x "Frequency" --y "Volume" --label "customer" --title "Customer Segmentation"
present /output/scatter-chart.svg

Important

• Always use CSV file input when data comes from table query. Do not manually copy values.
• Output is SVG, saved to /output/ by default.
• Always call present after generating a chart to display it inline.
• X-axis labels rotate automatically when there are many categories.