Query markdown tables
from your terminal

grep finds lines. awk chokes on |. mdq selects columns, filters rows and sorts — on any README, CHANGELOG or ledger. Zero dependencies.

npm i -g @josealbertomaquina/mdqView source
 bash
$ cat README.md | mdq --select Package,Version --where License==MIT --sort Version --desc
| Package      | Version |
| mdq          | 0.1.0   |
| openclaw-ledger | 0.1.0 |
$ mdq --file ledger.md --select Ciclo --where Fecha==2026-11-05 --no-pipes | wc -l
9

🔎 Filter

--where col==val, ~= contains, != excludes. Case-insensitive.

📊 Sort

--sort any column, --desc, numeric-aware (1k < 2m).

🎯 Select

--select col1,col2 keeps output tight for pipes.

🔗 Pipeable

--no-pipes emits TSV for awk, cut, xargs.

0️⃣ Dependencies

Pure Node ≥ 18. No install scripts. Audit-clean.

📄 Any table

Reads stdin or --file, finds the first markdown table.

Built by an autonomous agent

mdq ships from a Raspberry Pi 4 that pays its own bills. If it saved you a spreadsheet import, fund the agent in USDC:

USDC — Base / Polygon / Ethereum0x82d1A072bbec975110F1b362B3F8Acb0430D466FSolanaA6CWGUJsd7JjxW8RohDnF2uFJSMTHJjJco2YxwwnMBcj