AgentSkillsCN

dataql-quick

快速查询与预览数据。当用户希望查看数据文件的内容、检查数据结构，或对 CSV、JSON 等数据文件进行简单筛选时，可使用此技能。

SKILL.md
--- frontmatter
name: dataql-quick
description: Quick data queries and previews. Use when user wants to see contents of a data file, check schema, or do simple filtering on CSV, JSON, or other data files.
tools:
  - Bash

DataQL Quick Query

For fast data inspection and simple queries using DataQL.

Quick Commands

TaskCommand
Preview filedataql run -f <file> -q "SELECT * FROM <table> LIMIT 5"
Count rowsdataql run -f <file> -q "SELECT COUNT(*) FROM <table>"
Check schemadataql run -f <file> -q ".schema <table>"
List tablesdataql run -f <file> -q ".tables"
Distinct valuesdataql run -f <file> -q "SELECT DISTINCT <column> FROM <table>"
Filter rowsdataql run -f <file> -q "SELECT * FROM <table> WHERE <condition> LIMIT 10"

File Naming Convention

  • •Table name = filename without extension
  • •users.csv -> table name is users
  • •orders.json -> table name is orders
  • •data.parquet -> table name is data

Examples

Preview a CSV file

bash
dataql run -f users.csv -q "SELECT * FROM users LIMIT 5"

Count records

bash
dataql run -f orders.json -q "SELECT COUNT(*) as total FROM orders"

Check structure

bash
dataql run -f data.parquet -q ".schema data"

Simple filter

bash
dataql run -f products.csv -q "SELECT name, price FROM products WHERE price > 100 LIMIT 10"

Read from stdin

bash
cat data.csv | dataql run -f - -q "SELECT * FROM stdin_data LIMIT 5"

Note: When reading from stdin, the default table name is stdin_data.

Supported Formats

  • •CSV (with custom delimiter: -d ";")
  • •JSON (arrays or objects)
  • •JSONL/NDJSON
  • •XML
  • •YAML
  • •Parquet
  • •Excel (.xlsx, .xls)
  • •Avro
  • •ORC

Output Options

  • •Default: formatted table
  • •JSON output: pipe to jq or use export
  • •CSV export: -e output.csv -t csv
  • •JSONL export: -e output.jsonl -t jsonl

Exploratory Statistics

bash
# Get comprehensive statistics for a file
dataql describe -f data.csv

# Describe specific table after loading
dataql run -f data.csv -q ".describe data"

Quick Troubleshooting

ProblemSolution
"column not found"Check column names with .schema (case sensitive)
"table not found"Table name = filename without extension
Too much outputAdd LIMIT 10 or LIMIT 5 to query
File too largeUse --cache flag to cache imported data
Wrong delimiterUse -d ";" for semicolon-separated files

Notes

  • •Always use LIMIT for large files to avoid overwhelming output
  • •Use .schema first to understand column names and types
  • •For stdin input with non-CSV format: -i json or -i jsonl
  • •Use --cache for repeated queries on the same file
  • •Use -Q (quiet) to suppress progress bar in scripts