Tables

Summarize rows

rows_summarize · Reads only · Markdown

Counts rows, and sums, averages, minimums and maximums of number columns, overall and per group, in one call.

Free*

*Free up to 1,000,000 reads a month for the whole account, then $0.50 per 10,000 reads.

How much work is in the workshop, per status?

Your agent callsrows_summarizeFree*

The example business is Halverson Cycle & Repair, a bike shop with a repair workshop in Portland, Oregon. The reply is the one the tool really gives.

With these arguments

{
  "table": "Workshop jobs",
  "columns": [
    "Price",
    "Due"
  ],
  "group_by": "Status"
}

What comes back

{
  "table": "Workshop jobs",
  "all": {
    "count": 3,
    "Price": {
      "sum": 10000,
      "avg": 3333.333333,
      "min": 180,
      "max": 9400
    },
    "Due": {
      "min": "2026-09-30",
      "max": "2026-12-01"
    }
  },
  "group_by": "Status",
  "groups": [
    {
      "value": "On the stand",
      "count": 1,
      "Price": {
        "sum": 180,
        "avg": 180,
        "min": 180,
        "max": 180
      },
      "Due": {
        "min": "2026-10-22",
        "max": "2026-10-22"
      }
    },
    {
      "value": "Picked up",
      "count": 1,
      "Price": {
        "sum": 420,
        "avg": 420,
        "min": 420,
        "max": 420
      },
      "Due": {
        "min": "2026-09-30",
        "max": "2026-09-30"
      }
    }
  ]
}

Use it when

Someone asks for totals: how much is on order, how many per status, the latest date.

For the rows themselves, use rows_query.

What it does

Count rows, and sum, average, min and max number columns (min and max of dates), for the rows matching filters, overall and per value of group_by: one call instead of paging through every row. Filters work as in rows_query.

What comes back

all: the count and each measured column's sum, avg, min and max (min and max for dates); groups: the same per value of group_by, most rows first. Every reply also carries cost_usd and balance_usd.

Inputs

Name	Type	What it is
`table`Required	string	The table's name (or its table_id). 1 to 80 characters.
`columns`	array of strings	Number or date columns to measure. Up to 10 items.
`group_by`	string	A column to group the rows by. Up to 60 characters.
`filters`	array of objects	Up to 20 items.
`filters[].column`Required	string	Up to 60 characters.
`filters[].op`	string	One of: eq,neq,contains,not_contains,gt,gte,lt,lte,empty,not_empty,in.
`filters[].value`	any
`match`	string	One of: all,any. Default: all.

The call

Over MCP, call rows_summarize on https://api.conecto.chat/mcp/tables. Over REST:

Terminal

curl -X POST https://api.conecto.chat/v1/tools/rows_summarize \
  -H "Authorization: Bearer $CONECTO_KEY" \
  -H "Content-Type: application/json" \
  -d '{"table":"Workshop jobs","columns":["Price","Due"],"group_by":"Status"}'

When it refuses

A refused call costs nothing and says what to do next.

Code	When	What to do
`not_found`404	No table has that name or table_id, or the table has no column of that name.	Call tables_list for the tables and their columns.
`invalid_input`400	A filter's op is unknown, compares with the wrong kind of value, or uses gt, gte, lt or lte on a column that is not a number or a date.	Use the ops this page lists, with values of the column's type.
`invalid_input`400	A measured column is not a number or date column.	Measure number or date columns; count the others with group_by.

More Tables tools

All 12 tools

Give this to your agent.

Set up Conecto from https://conecto.chat/SKILL.md

Summarize rows

An example

Use it when

What it does

What comes back

Inputs

The call

When it refuses

More Tables tools

Give this to your agent.