Skip to content

Data Analysis with DuckDB

Lets the agent inspect, query, summarize and export Excel and CSV files using SQL.

Stars
84,876
Licence
MIT
Files
3 files

What it does

When you upload Excel or CSV files, this skill tells the agent to use a small Python script that loads the spreadsheets into DuckDB, a fast local SQL engine. The agent can then look at the column structure, run SQL queries such as totals, groupings, filters and joins across files, produce statistical summaries, and save results as CSV, JSON or Markdown. It is a way to get spreadsheet analysis done with real queries rather than manual inspection.

Great for

  • Analysts who want quick SQL-based answers from spreadsheet exports
  • People combining several Excel or CSV files with joins
  • Users who need summary statistics or filtered data exported to a file

Skip it if

  • Very sensitive datasets sent to an agent environment, since the data is processed and cached locally in temp storage and the skill's behaviour depends on the host environment
  • Live database work or files outside the expected uploads folder, because the instructions assume /mnt/user-data paths

Try asking

  • “Inspect this Excel workbook and tell me what sheets and columns it has”
  • “Show total revenue by category and month from my Sales sheet”
  • “Join sales.csv with customers.csv and list the top 10 customers by spend”
  • “Export all orders over 1000 to a CSV file in my outputs folder”
Step-by-step guide: how to use Data Analysis with DuckDBInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 3 original files, the MIT licence and an attribution note.

    data-analysis.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip data-analysis.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip data-analysis.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Run inspect first so the agent uses the correct SQL table names, which may be sanitized or renamed on collisions.
  • Install duckdb and openpyxl in the same Python interpreter the agent uses, since the script exits with an instruction if they are missing.
  • Use a unique output filename, because existing files at the target path are overwritten without warning.

What's inside

  • AGENTS.md
  • scripts/analyze.py
  • SKILL.md

Redistributed unmodified from bytedance/deer-flow under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.