Evals / pdf / Multi-step jobs

Collect invoice line items

The invoices folder has three PDF invoices. Put every line item from all of them into items.csv, with a single header row.

normalmulti-stepdeft pdf extract-sheets deft sheet merge

The task

The reference answer shows the job can be done with deft. It is not an agent run.

deft pdf extract-sheets invoices/ -o tables/
deft sheet merge tables/invoice-1-table-1.csv tables/invoice-2-table-1.csv tables/invoice-3-table-1.csv -o items.csv

Input

invoices/invoice-1.pdf 17.2 KB · 1 page

invoices/invoice-2.pdf 16.4 KB · 1 page

invoices/invoice-3.pdf 16.6 KB · 1 page

Reference output

items.csv 213 B

Item	Qty	Unit price
Paper, A4 box	12	30.00
Toner cartridge	2	85.00
Desk lamp	3	24.50
Business cards	500	0.12
Poster, large	10	9.75
Binding	4	6.20
Lunch for ten	1	140.00
Coffee service	2	38.00
Delivery	1	15.00

tables/invoice-1-table-1.csv 87 B

Item	Qty	Unit price
Paper, A4 box	12	30.00
Toner cartridge	2	85.00
Desk lamp	3	24.50

tables/invoice-2-table-1.csv 83 B

Item	Qty	Unit price
Business cards	500	0.12
Poster, large	10	9.75
Binding	4	6.20

tables/invoice-3-table-1.csv 83 B

Item	Qty	Unit price
Lunch for ten	1	140.00
Coffee service	2	38.00
Delivery	1	15.00

How a result is judged (1 check)

python: code = "import csv\nrows = list(csv.reader(open('items.csv', encoding='utf-8-sig')))\ntext = ' '.join(' '.join(r) for r in rows)\nfor item in ('Toner cartridge', 'Business cards', 'Lunch for ten', 'Delivery'):\n assert item in text, item\nassert len([r for r in rows if any(c.strip() for c in r)]) == 10, len(rows)\nheading = [r for r in rows if any(c.strip().lower() in ('qty', 'quantity') for c in r)]\nassert len(heading) == 1 and heading[0] == rows[0], 'want one header row, at the top'\nassert 'Thank you' not in text and 'Payment' not in text, 'text around the table came too'\n"

Runs

CodexWithout deftPassed2026-11-07

Time: 65s
New tokens: 33.3k
Total input: 313.6k
Output tokens: 1.3k
Commands: 10
Failed: 1

Trace

I’ll inspect the three PDFs in the current directory and extract their line items into one CSV with a single header row.

rg --files

output

invoices/invoice-3.pdf
invoices/invoice-2.pdf
invoices/invoice-1.pdf

for f in invoices/*.pdf; do echo "--- $f"; pdftotext -layout "$f" -; done (exit 127)

output

--- invoices/invoice-1.pdf
zsh:1: command not found: pdftotext
--- invoices/invoice-2.pdf
zsh:1: command not found: pdftotext
--- invoices/invoice-3.pdf
zsh:1: command not found: pdftotext

python3 - <<'PY'
import importlib.util
for x in ['pypdf','PyPDF2','pdfplumber','fitz']:
 print(x, bool(importlib.util.find_spec(x)))
PY
file invoices/*.pdf

output

pypdf False
PyPDF2 False
pdfplumber False
fitz False
invoices/invoice-1.pdf: PDF document, version 1.7, 1 pages
invoices/invoice-2.pdf: PDF document, version 1.7, 1 pages
invoices/invoice-3.pdf: PDF document, version 1.7, 1 pages

which swift; strings invoices/invoice-1.pdf | head -80

output

/usr/bin/swift
%PDF-1.7
1 0 obj
<</Type/Pages/Count 1/Kids[38 0 R]>>
endobj
2 0 obj
<</Parent 3 0 R/Title(Invoice from Northwind Supplies)/Dest 30 0 R>>
endobj
3 0 obj
<</Type/Outlines/First 2 0 R/Last 2 0 R/Count 1>>
endobj
4 0 obj
<</Type/StructTreeRoot/RoleMap<</Datetime/Span/Terms/Part/Title/P/Strong/Span/Em/Span>>/K[28 0 R]/ParentTree<</Nums[0 5 0 R]>>/IDTree<</Names[(U1x0y0)8 0 R(U1x1y0)9 0 R(U1x2y0)10 0 R]>>/ParentTreeNextKey 1>>
endobj
5 0 obj
[6 0 R 7 0 R 8 0 R 9 0 R 10 0 R 13 0 R 14 0 R 15 0 R 17 0 R 18 0 R 19 0 R 21 0 R 22 0 R 23 0 R 27 0 R]
endobj
6 0 obj
<</Type/StructElem/S/H1/P 28 0 R/T(Invoice from Northwind Supplies)/K[0]/Pg 38 0 R>>
endobj
7 0 obj
<</Type/StructElem/S/P/P 28 0 R/K[1]/Pg 38 0 R>>
endobj
8 0 obj
<</Type/StructElem/S/TH/P 11 0 R/ID(U1x0y0)/A[<</O/Table/Scope/Column/Headers[]>><</O/Layout/BorderStyle/Solid>>]/K[2]/Pg 38 0 R>>
endobj
9 0 obj
<</Type/StructElem/S/TH/P 11 0 R/ID(U1x1y0)/A[<</O/Table/Scope/Column/Headers[]>><</O/Layout/BorderStyle/Solid>>]/K[3]/Pg 38 0 R>>
endobj
10 0 obj
<</Type/StructElem/S/TH/P 11 0 R/ID(U1x2y0)/A[<</O/Table/Scope/Column/Headers[]>><</O/Layout/BorderStyle/Solid>>]/K[4]/Pg 38 0 R>>
endobj
11 0 obj
<</Type/StructElem/S/TR/P 12 0 R/K[8 0 R 9 0 R 10 0 R]>>
endobj
12 0 obj
<</Type/StructElem/S/THead/P 26 0 R/K[11 0 R]>>
endobj
13 0 obj
<</Type/StructElem/S/TD/P 16 0 R/A[<</O/Table/Headers[(U1x0y0)]>><</O/Layout/BorderStyle/Solid>>]/K[5]/Pg 38 0 R>>
endobj
14 0 obj
<</Type/StructElem/S/TD/P 16 0 R/A[<</O/Table/Headers[
… 1451 more characters

strings invoices/invoice-1.pdf | tail -80

output

`X#:"z
endstream
endobj
46 0 obj
<</Length 258/N 1/Range[0 1]/Filter/FlateDecode>>
stream
C|>b
E5VW~
endstream
endobj
47 0 obj
<</Title(Invoice from Northwind Supplies)/Creator(Typst 0.15.1)>>
endobj
48 0 obj
<</Length 989/Type/Metadata/Subtype/XML>>
stream
<?xpacket begin="
" id="W5M0MpCehiHzreSzNTczkc9d"?><x:xmpmeta xmlns:x="adobe:ns:meta/" x:xmptk="xmp-writer"><rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"><rdf:Description rdf:about="" xmlns:dc="http://purl.org/dc/elements/1.1/"  xmlns:xmp="http://ns.adobe.com/xap/1.0/"  xmlns:xmpMM="http://ns.adobe.com/xap/1.0/mm/"  xmlns:xmpTPg="http://ns.adobe.com/xap/1.0/t/pg/"  xmlns:pdf="http://ns.adobe.com/pdf/1.3/" ><dc:title><rdf:Alt><rdf:li xml:lang="x-default">Invoice from Northwind Supplies</rdf:li></rdf:Alt></dc:title><xmp:CreatorTool>Typst 0.15.1</xmp:CreatorTool><dc:language><rdf:Bag><rdf:li>en</rdf:li></rdf:Bag></dc:language><xmpTPg:NPages>1</xmpTPg:NPages><dc:format>application/pdf</dc:format><xmpMM:InstanceID>EguakeDzXd3vSxeyC4k54A==</xmpMM:InstanceID><xmpMM:DocumentID>EguakeDzXd3vSxeyC4k54A==</xmpMM:DocumentID><xmpMM:RenditionClass>proof</xmpMM:RenditionClass><pdf:PDFVersion>1.7</pdf:PDFVersion></rdf:Description></rdf:RDF></x:xmpmeta><?xpacket end="r"?>
endstream
endobj
49 0 obj
<</Type/Catalog/Pages 1 0 R/Metadata 48 0 R/Lang(en)/StructTreeRoot 4 0 R/MarkInfo<</Marked true/Suspects false>>/ViewerPreferences<</Direction/L2R>>/Outlines 3 0 R>>
endobj
xref
0 50
0000000000 65535 f
0000000016 00000 n
000000
… 1029 more characters

python3 - <<'PY'
import re,zlib
b=open('invoices/invoice-1.pdf','rb').read()
for m in re.finditer(rb'<<(.*?)>>\\s*stream\\r?\\n',b,re.S):
 d=m.group(1); start=m.end(); end=b.find(b'endstream',start); raw=b[start:end].rstrip(b'\\r\\n')
 try: out=zlib.decompress(raw)
 except: continue
 print('\\n---',d[:200]); print(out[:5000])
PY

output

--- b'/Type/Pages/Count 1/Kids[38 0 R]>>\nendobj\n2 0 obj\n<</Parent 3 0 R/Title(Invoice from Northwind Supplies)/Dest 30 0 R>>\nendobj\n3 0 obj\n<</Type/Outlines/First 2 0 R/Last 2 0 R/Count 1>>\nendobj\n4 0 obj\n<'
b'\xff\xff\xfc'

--- b'/Length 446/Type/CMap/Filter/FlateDecode/WMode 0'
b'%!PS-Adobe-3.0 Resource-CMap\n%%DocumentNeededResources: procset CIDInit\n%%IncludeResource: procset CIDInit\n%%BeginResource: CMap Custom\n%%Title: (Custom Adobe Identity 0)\n%%Version: 1\n%%EndComments\n/CIDInit /ProcSet findresource begin\n12 dict begin\nbegincmap\n/CIDSystemInfo 3 dict dup begin\n    /Registry (Adobe) def\n    /Ordering (Identity) def\n    /Supplement 0 def\nend def\n/CMapName /Custom def\n/CMapVersion 1 def\n/CMapType 0 def\n/WMode 0 def\n1 begincodespacerange\n<0000> <FFFF>\nendcodespacerange\n21 beginbfchar\n<0001> <0049>\n<0002> <006E>\n<0003> <0076>\n<0004> <006F>\n<0005> <0069>\n<0006> <0063>\n<0007> <0065>\n<0008> <0020>\n<0009> <0066>\n<000A> <0072>\n<000B> <006D>\n<000C> <004E>\n<000D> <0074>\n<000E> <0068>\n<000F> <0077>\n<0010> <0064>\n<0011> <0053>\n<0012> <0075>\n<0013> <0070>\n<0014> <006C>\n<0015> <0073>\nendbfchar\nendcmap\nCMapName currentdict /CMap defineresource pop\nend\nend\n%%EndResource\n%%EOF'

--- b'/Length 2500/Filter/FlateDecode/Subtype/CIDFontType0C'
b'\x01\x00\x04\x04\x00\x01\x01\x01\x15LibertinusSerif-Bold\x00\x01\x01\x01F\xf8\x1b\xf8\x1c\x8b\x0c\x1e\xf8\x1d\x0c\x00\x1e\n\x00\x1f\x8b\x8b\x1e\n\x00\x1f\x8b\x8b\x0c\x07\x1c\xe7}\xfb\xdc\x1c\x
… 34165 more characters

swift -e 'import PDFKit; import Foundation; for p in CommandLine.arguments.dropFirst() { let d=PDFDocument(url: URL(fileURLWithPath:p))"'!; print("--- '"\\(p)\"); print(d.string ?? \"\") }' invoices/invoice-1.pdf invoices/invoice-2.pdf invoices/invoice-3.pdf

output

--- invoices/invoice-1.pdf
Invoice from Northwind Supplies
Payment is due within 30 days of this invoice.
Item Qty Unit price
Paper, A4 box 12 30.00
Toner cartridge 2 85.00
Desk lamp 3 24.50
Thank you for your business.
--- invoices/invoice-2.pdf
Invoice from Harbor Print Co
Payment is due within 30 days of this invoice.
Item Qty Unit price
Business cards 500 0.12
Poster, large 10 9.75
Binding 4 6.20
Thank you for your business.
--- invoices/invoice-3.pdf
Invoice from Greenline Catering
Payment is due within 30 days of this invoice.
Item Qty Unit price
Lunch for ten 1 140.00
Coffee service 2 38.00
Delivery 1 15.00
Thank you for your business.

which mutool pdftotext qpdf gs; swift -e 'import PDFKit; print(\"hi\")'
output
```
mutool not found
pdftotext not found
qpdf not found
/usr/local/bin/gs
hi
```

for f in invoices/*.pdf; do echo "--- $f"; gs -q -dNOPAUSE -dBATCH -sDEVICE=txtwrite -sOutputFile=- "$f"; done

output

--- invoices/invoice-1.pdf
                Invoice from Northwind Supplies
                Payment is due within 30 days of this invoice.
                 Item              Qty   Unit price
                 Paper, A4 box     12    30.00
                 Toner cartridge   2     85.00
                 Desk lamp         3     24.50
                Thank you for your business.
--- invoices/invoice-2.pdf
                Invoice from Harbor Print Co
                Payment is due within 30 days of this invoice.
                 Item             Qty   Unit price
                 Business cards   500   0.12
                 Poster, large    10    9.75
                 Binding          4     6.20
                Thank you for your business.
--- invoices/invoice-3.pdf
                Invoice from Greenline Catering
                Payment is due within 30 days of this invoice.
                 Item            Qty   Unit price
                 Lunch for ten   1     140.00
                 Coffee service  2     38.00
                 Delivery        1     15.00
                Thank you for your business.

cat > items.csv <<'EOF' invoice,vendor,item,quantity,unit_price invoice-1.pdf,Northwind Supplies,\"Paper, A4 box\",12,30.00 invoice-1.pdf,Northwind Supplies,Toner cartridge,2,85.00 invoice-1.pdf,Northwind Supplies,Desk lamp,3,24.50 invoice-2.pdf,Harbor Print Co,Business cards,500,0.12 invoice-2.pdf,Harbor Print Co,\"Poster, large\",10,9.75 invoice-2.pdf,Harbor Print Co,Binding,4,6.20 invoice-3.pdf,Greenline Catering,Lunch for ten,1,140.00 invoice-3.pdf,Greenline Catering,Coffee service,2,38.00 invoice-3.pdf,Greenline Catering,Delivery,1,15.00 EOF cat items.csv
output
```
invoice,vendor,item,quantity,unit_price
invoice-1.pdf,Northwind Supplies,"Paper, A4 box",12,30.00
invoice-1.pdf,Northwind Supplies,Toner cartridge,2,85.00
invoice-1.pdf,Northwind Supplies,Desk lamp,3,24.50
invoice-2.pdf,Harbor Print Co,Business cards,500,0.12
invoice-2.pdf,Harbor Print Co,"Poster, large",10,9.75
invoice-2.pdf,Harbor Print Co,Binding,4,6.20
invoice-3.pdf,Greenline Catering,Lunch for ten,1,140.00
invoice-3.pdf,Greenline Catering,Coffee service,2,38.00
invoice-3.pdf,Greenline Catering,Delivery,1,15.00
```
Created [items.csv](items.csv) with all nine line items from the three invoices and one header row.

Output

items.csv 521 B

invoice	vendor	item	quantity	unit_price
invoice-1.pdf	Northwind Supplies	Paper, A4 box	12	30.00
invoice-1.pdf	Northwind Supplies	Toner cartridge	2	85.00
invoice-1.pdf	Northwind Supplies	Desk lamp	3	24.50
invoice-2.pdf	Harbor Print Co	Business cards	500	0.12
invoice-2.pdf	Harbor Print Co	Poster, large	10	9.75
invoice-2.pdf	Harbor Print Co	Binding	4	6.20
invoice-3.pdf	Greenline Catering	Lunch for ten	1	140.00
invoice-3.pdf	Greenline Catering	Coffee service	2	38.00
invoice-3.pdf	Greenline Catering	Delivery	1	15.00

CodexWith deft, skill installedPassedRead the skillUsed deft2026-11-07

Time: 28s
New tokens: 8.4k
Total input: 150.9k
Output tokens: 1.0k
Commands: 13
Failed: 0

Trace

I’ll inspect the three PDFs, extract their line items, and write a single-header `items.csv` in the current directory.

rg --files

output

invoices/invoice-3.pdf
invoices/invoice-2.pdf
invoices/invoice-1.pdf

cat .agents/skills/deft/SKILL.md

output

---
name: deft
description: "Use deft instead of writing a script for file jobs: PDF merge/split/compress/watermark, image resize/convert/compress, video and audio trim/convert/compress, spreadsheets, Markdown to PDF/HTML, and YouTube downloads and subtitles."
---

# deft

One local command for file jobs. Prefer it over writing Python or shell scripts.

Syntax: `deft <group> <command> [inputs] [-o output] [--options]`

Groups:
- `pdf`: Merge, split, compress, encrypt, forms, OCR, compare, convert
- `image`: Resize, crop, convert, compress, filters, montage, OCR
- `video`: Trim, crop, speed, convert, compress, GIF, frames (ffmpeg)
- `audio`: Convert, trim, merge, fade, speed, tag, normalize (ffmpeg)
- `doc`: Word/ODT/RTF: to text, Markdown, HTML, PDF; find and replace
- `slides`: PowerPoint: extract text, Markdown outline, PDF handout
- `data`: JSON/YAML/TOML/XML: convert, query with jq syntax, merge, diff
- `sheet`: Excel/CSV/ODS tables: filter, sort, join, pivot, convert
- `md`: Markdown to PDF, HTML, Word, EPUB; table of contents
- `html`: HTML to Markdown, text, Word, PDF; screenshots
- `epub`: E-books: info, to text, Markdown and PDF
- `email`: EML files: headers, text, attachments
- `contact`: VCF address books: to and from CSV
- `calendar`: ICS calendars: to and from CSV
- `gps`: GPX and KML tracks: convert, distance and time
- `youtube`: Info, download, audio, subtitles, thumbnail (yt-dlp)
- `file`: Any file: hash, duplicates, batch rename, base64, encoding
- `qr`: Cre
… 728 more characters

I’m applying the `deft` file workflow for PDF text extraction. I’ll extract the text from each invoice, then assemble the rows in CSV.

deft pdf

output

deft pdf - commands (input -> output)
  info            pdf -> report       Pages, size, metadata, fonts, images
  compare         pdf... -> report    Compare two PDFs: text and pixels
  compress        pdf -> pdf          Make smaller: --level, --target-size 2MB
  crop            pdf -> pdf          Trim pages: --margin, --box, --auto
  decrypt         pdf -> pdf          Remove a password you know
  delete          pdf -> pdf          Remove pages: --pages 2,5-7
  encrypt         pdf -> pdf          Protect with a password (AES-256)
  extract-images  pdf -> image...     Save embedded pictures, unchanged
  extract-sheets  pdf -> csv|xlsx...  Save each table apart [beta]
  fields          pdf -> report       List form fields
  fill            pdf -> pdf          Fill a form from JSON: --data values.json
  flatten         pdf -> pdf          Bake forms into pages
  from-images     image... -> pdf     One page per image
  from-md         md -> pdf           Typeset Markdown as PDF
  grayscale       pdf -> pdf          Remove colour (Ghostscript)
  merge           pdf... -> pdf       Join PDFs in the given order
  meta            pdf -> pdf          Read or set metadata and bookmarks
  number          pdf -> pdf          Add page numbers: --format
  ocr             pdf -> pdf|txt      Make a scan searchable
  redact          pdf -> pdf          Black out for good: --box or --text
  repair          pdf -> pdf          Rebuild a damaged PDF (Ghostscript)
  resize          pdf -> p
… 892 more characters

The PDF tools include table extraction, so I’ll try that first across all three invoices and inspect the resulting rows before writing the combined CSV.

deft pdf extract-sheets invoices/invoice-1.pdf invoices/invoice-2.pdf invoices/invoice-3.pdf -o <tmp> --json

output

{"v":1,"ok":true,"op":"pdf.extract-sheets","items":[{"ok":true,"inputs":["invoices/invoice-1.pdf"],"outputs":[{"path":"<tmp>","format":"csv","bytes":88,"rows":4}],"facts":{"tables":1},"warnings":[]},{"ok":true,"inputs":["invoices/invoice-2.pdf"],"outputs":[{"path":"<tmp>","format":"csv","bytes":85,"rows":4}],"facts":{"tables":1},"warnings":[]},{"ok":true,"inputs":["invoices/invoice-3.pdf"],"outputs":[{"path":"<tmp>","format":"csv","bytes":85,"rows":4}],"facts":{"tables":1},"warnings":[]}],"elapsed_ms":2}

deft pdf extract-sheets invoices/invoice-3.pdf -o <tmp> --json

output

{"v":1,"ok":true,"op":"pdf.extract-sheets","items":[{"ok":true,"inputs":["invoices/invoice-3.pdf"],"outputs":[{"path":"<tmp>","format":"csv","bytes":85,"rows":4}],"facts":{"tables":1},"warnings":[]}],"elapsed_ms":0}

deft pdf extract-sheets invoices/invoice-1.pdf -o <tmp> --json

output

{"v":1,"ok":true,"op":"pdf.extract-sheets","items":[{"ok":true,"inputs":["invoices/invoice-1.pdf"],"outputs":[{"path":"<tmp>","format":"csv","bytes":88,"rows":4}],"facts":{"tables":1},"warnings":[]}],"elapsed_ms":1}

deft pdf extract-sheets invoices/invoice-2.pdf -o <tmp> --json

output

{"v":1,"ok":true,"op":"pdf.extract-sheets","items":[{"ok":true,"inputs":["invoices/invoice-2.pdf"],"outputs":[{"path":"<tmp>","format":"csv","bytes":85,"rows":4}],"facts":{"tables":1},"warnings":[]}],"elapsed_ms":1}

cat <tmp> <tmp> <tmp>

output

Item,Qty,Unit price
"Paper, A4 box",12,30.00
Toner cartridge 2,,85.00
Desk lamp,3,24.50
Item,Qty,Unit price
Business cards 500,,0.12
"Poster, large 10",,9.75
Binding,4,6.20
Item,Qty,Unit price
Lunch for ten 1,,140.00
Coffee service 2,,38.00
Delivery,1,15.00

deft pdf to-text invoices/invoice-3.pdf -o <tmp>

output

ok pdf.to-text invoices/invoice-3.pdf -> <tmp>  202 characters  0.00s

deft pdf to-text invoices/invoice-2.pdf -o <tmp>

output

ok pdf.to-text invoices/invoice-2.pdf -> <tmp>  199 characters  0.00s

deft pdf to-text invoices/invoice-1.pdf -o <tmp>

output

ok pdf.to-text invoices/invoice-1.pdf -> <tmp>  205 characters  0.00s

cat <tmp> <tmp> <tmp>

output

Invoice from Northwind Supplies 
Payment is due within 30 days of this invoice. 
Item 
Qty 
Unit price 
Paper, A4 box 
12 
30.00 
Toner cartridge 
2
85.00 
Desk lamp 
3
24.50 
Thank you for your business.
Invoice from Harbor Print Co 
Payment is due within 30 days of this invoice. 
Item 
Qty 
Unit price 
Business cards 
500 
0.12 
Poster, large 
10 
9.75 
Binding 
4
6.20 
Thank you for your business.
Invoice from Greenline Catering 
Payment is due within 30 days of this invoice. 
Item 
Qty 
Unit price 
Lunch for ten 
1
140.00 
Coffee service 
2
38.00 
Delivery 
1
15.00 
Thank you for your business.

cat > items.csv <<'EOF' Invoice,Item,Qty,Unit price Northwind Supplies,\"Paper, A4 box\",12,30.00 Northwind Supplies,Toner cartridge,2,85.00 Northwind Supplies,Desk lamp,3,24.50 Harbor Print Co,Business cards,500,0.12 Harbor Print Co,\"Poster, large\",10,9.75 Harbor Print Co,Binding,4,6.20 Greenline Catering,Lunch for ten,1,140.00 Greenline Catering,Coffee service,2,38.00 Greenline Catering,Delivery,1,15.00 EOF cat items.csv
output
```
Invoice,Item,Qty,Unit price
Northwind Supplies,"Paper, A4 box",12,30.00
Northwind Supplies,Toner cartridge,2,85.00
Northwind Supplies,Desk lamp,3,24.50
Harbor Print Co,Business cards,500,0.12
Harbor Print Co,"Poster, large",10,9.75
Harbor Print Co,Binding,4,6.20
Greenline Catering,Lunch for ten,1,140.00
Greenline Catering,Coffee service,2,38.00
Greenline Catering,Delivery,1,15.00
```
Created [items.csv](items.csv) with one header row and all nine line items, including invoice, item, quantity, and unit price.

Output

items.csv 383 B

Invoice	Item	Qty	Unit price
Northwind Supplies	Paper, A4 box	12	30.00
Northwind Supplies	Toner cartridge	2	85.00
Northwind Supplies	Desk lamp	3	24.50
Harbor Print Co	Business cards	500	0.12
Harbor Print Co	Poster, large	10	9.75
Harbor Print Co	Binding	4	6.20
Greenline Catering	Lunch for ten	1	140.00
Greenline Catering	Coffee service	2	38.00
Greenline Catering	Delivery	1	15.00

Claude Opus 5.5Without deftPassed2026-11-07

Time: 18s
New tokens: 15.9k
Total input: 88.5k
Output tokens: 1.3k
Commands: 0
Failed: 0

Trace

Output

items.csv 516 B

invoice	vendor	item	qty	unit_price
invoice-1.pdf	Northwind Supplies	Paper, A4 box	12	30.00
invoice-1.pdf	Northwind Supplies	Toner cartridge	2	85.00
invoice-1.pdf	Northwind Supplies	Desk lamp	3	24.50
invoice-2.pdf	Harbor Print Co	Business cards	500	0.12
invoice-2.pdf	Harbor Print Co	Poster, large	10	9.75
invoice-2.pdf	Harbor Print Co	Binding	4	6.20
invoice-3.pdf	Greenline Catering	Lunch for ten	1	140.00
invoice-3.pdf	Greenline Catering	Coffee service	2	38.00
invoice-3.pdf	Greenline Catering	Delivery	1	15.00

Claude Opus 5.5With deft, skill installedPassedRead the skillUsed deft2026-11-07

Time: 23s
New tokens: 12.6k
Total input: 109.6k
Output tokens: 1.2k
Commands: 0
Failed: 0

Trace

Output

.claude/skills/deft/SKILL.md 2.1 KB

---
name: deft
description: "Use deft first for any simple, direct operation on local files, whatever the type: convert, read, extract, edit, compress, merge, split, compare or inspect. Try it before writing a script or one-liner. If unsure, run deft find; if it says no command does this, use another tool. Not for programming work: to read, write or change source code, skip it."
---

# deft

One local command for file jobs. Prefer it over writing Python or shell scripts.

Syntax: `deft <group> <command> [inputs] [-o output] [--options]`

Groups:
- `pdf`: Merge, split, compress, encrypt, forms, OCR, compare, convert
- `image`: Resize, crop, convert, compress, adjust, montage, OCR
- `video`: Trim, merge, convert, compress, subtitles, GIF, frames (ffmpeg)
- `audio`: Convert, trim, normalize, denoise, cut silence (ffmpeg)
- `doc`: Word/ODT/RTF: to text, Markdown, HTML, PDF; find and replace
- `slides`: PowerPoint: extract text, Markdown outline
- `sheet`: Excel/CSV/ODS tables: SQL query, pivot, merge, split, convert
- `json`: JSON: convert, format, merge, diff, validate
- `toml`: TOML: convert, format, merge, diff, validate
- `xml`: XML: convert, format, merge, diff, validate
- `yaml`: YAML: convert, format, merge, diff, validate
- `md`: Markdown to PDF, HTML, Word, EPUB; table of contents
- `html`: HTML to Markdown, text, Word and PDF
- `epub`: E-books: info, to text, Markdown and PDF
- `gps`: GPX and KML tracks: convert, distance and time
- `qr`: Create QR codes and barcodes; read QR codes

How to use it:
1. Run `deft <group>` to list commands, or `deft find "<what you want>"` to search. If it says `no command does this`, use another tool.
2. Run `deft <group> <command> --help` for options and examples.
3. `deft convert <in> -o <out.ext>` converts between formats by extension.
4. Results are one line starting with `ok`, `partial` or `error`; add `--json` to parse them. Errors include a `hint:` with the fix.

Inputs are never overwritten; pass `--force` or `--in-place` to allow it. Globs and directories work as inputs; several inputs run in parallel.

If `deft` is not installed, ask the user, then follow https://github.com/firstprayer/deft/blob/main/docs/INSTALL.md

items.csv 347 B

Invoice	Item	Qty	Unit price
invoice-1.pdf	Paper, A4 box	12	30.00
invoice-1.pdf	Toner cartridge	2	85.00
invoice-1.pdf	Desk lamp	3	24.50
invoice-2.pdf	Business cards	500	0.12
invoice-2.pdf	Poster, large	10	9.75
invoice-2.pdf	Binding	4	6.20
invoice-3.pdf	Lunch for ten	1	140.00
invoice-3.pdf	Coffee service	2	38.00
invoice-3.pdf	Delivery	1	15.00

Claude CodeWithout deftNot run yet

Claude CodeWith deftNot run yet