table

name	table
description	Query, filter, aggregate, group, sort, and transform CSV/XLSX files. Always use instead of awk, python, or manual parsing for tabular data.
metadata	{"nessi":{"command":"table","enabled":true}}

Table

Work with CSV and XLSX files: inspect structure, query with filters, aggregations, and calculations, and export results.

Inspect

table info /input/data.xlsx
table columns /input/data.xlsx --sheet "Sales"
table peek /input/data.xlsx --rows 10 --columns "name,revenue"

Query

The query command combines filtering, aggregation, calculated columns, projection, aliases, sorting, and limiting in one step.

Filter rows

table query /input/sales.xlsx --where "year = 2024" --output /output/filtered.csv
table query /input/sales.xlsx --where "status = active" --where "amount > 100" --output /output/result.csv

Aggregate with grouping

table query /input/sales.xlsx \
  --select "region, sum(revenue) as Total, count() as Deals, avg(price) as Avg" \
  --group "region" \
  --output /output/by-region.csv

Aggregation functions: count(), sum(col), avg(col), min(col), max(col), median(col).

Calculated columns with calc()

Use calc(expression) in --select to compute new columns. Supports +, -, *, / and parentheses.

Row-level calculations (like Excel formulas — applied to every row):

table query /input/products.xlsx \
  --select "product, price, calc(price * 1.19) as BruttoPreis" \
  --output /output/with-tax.csv

table query /input/orders.xlsx \
  --select "item, quantity, unit_price, calc(quantity * unit_price) as Total" \
  --output /output/totals.csv

Aggregation-level calculations (combine multiple aggregations):

table query /input/sales.xlsx \
  --select "region, sum(revenue) as Rev, sum(cost) as Cost, calc(sum(revenue) - sum(cost)) as Profit" \
  --group "region" \
  --output /output/profit.csv

table query /input/data.xlsx \
  --select "category, calc(sum(revenue) / count()) as AvgRevenue, calc((sum(revenue) - sum(cost)) / sum(revenue) * 100) as MarginPct" \
  --group "category" \
  --output /output/margins.csv

Project and rename columns

table query /input/data.xlsx --select "product as Produkt, revenue as Umsatz"

Sort and limit

table query /input/sales.xlsx \
  --select "product, sum(revenue) as Revenue" \
  --group "product" \
  --sort "Revenue desc" \
  --limit 5 \
  --output /output/top5.csv

Full example

table query /input/sales.xlsx \
  --select "region, sum(revenue) as Revenue, count() as Orders, calc(sum(revenue) / count()) as AvgOrder" \
  --where "year >= 2023" \
  --group "region" \
  --sort "Revenue desc" \
  --limit 10 \
  --output /output/by-region.csv

Pipeline: table → chart

Query results feed directly into chart commands:

table query /input/sales.xlsx \
  --select "region as Region, sum(revenue) as Revenue" \
  --group "region" --output /output/agg.csv

chart bar /output/agg.csv --x "Region" --y "Revenue" --title "Revenue by Region"
present /output/bar-chart.svg

Export & Transform

table export /input/data.xlsx --output /output/data.csv
table append /input/data.csv --json '[{"name":"Alice","age":"30"}]' --output /output/appended.csv
table replace /input/data.csv --column "status" --old "pending" --new "done" --output /output/replaced.csv

Filter operators

Operator	Example	Notes
= / !=	status = active	Case-insensitive. = (empty value) matches empty cells
is empty	Einlieferer is empty	Matches rows where the column is blank
is not empty	email is not empty	Matches rows where the column has a value
> < >= <=	amount > 100	Auto-detects numbers and dates
contains	name contains Smith	Substring match
starts_with	email starts_with admin	Prefix match
matches	email matches ^.*@gmail\.com$	Regex (case-insensitive)

Output behavior

• With --output /output/file.csv → writes CSV to file
• Without --output → prints result directly to stdout (useful for quick counts and inspections)

Important

• Do not use awk, node -e, python, grep, or manual CSV parsing. This skill handles filtering, aggregation, calculations, counting, grouping, sorting, and export.
• For XLSX with multiple sheets, use --sheet "SheetName".
• Pipe query output directly into chart for visualization — no manual value extraction needed.
• Always show table results via CSV + present, not via the card table layout. Write to --output /output/result.csv, then call present /output/result.csv. This avoids manually copying data into tool arguments and handles any number of rows. The card tool is only for small, hand-crafted data — table query results should always flow through files.