Evals / documents / Read a document

Read a deck with notes

Extract all the text of show.pptx, including the speaker notes, into show.txt.

normalsingledeft slides to-text

The task

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

deft slides to-text show.pptx --notes -o show.txt

Input

show.pptx 38.2 KB
Rendered by macOS Quick Look, first page

Rendered by macOS Quick Look, first page

Text inside
[ppt/slides/slide1.xml] Title of the first slide Subtitle of the first slide This bit is in italic green
[ppt/slides/slide2.xml] This is the second slide It has bullet points on it They’re fun, aren’t they? Especially in a different font like Arial Black at 16 point!

Reference output

show.txt 326 B
Title of the first slide

Subtitle of the first slide

This bit is in italic green


I am the notes of the first slide



This is the second slide

It has bullet points on it

They’re fun, aren’t they?

Especially in a different font like Arial Black at 16 point!


These are the notes of the 2nd slide
THIS LINE IS BOLD
How a result is judged (1 check)
  • text: file = "show.txt", contains = ["Title of the first slide", "I am the notes of the first slide"]

Runs

CodexWithout deftPassed2026-11-07

Time
19s
New tokens
5.8k
Total input
70.1k
Output tokens
0.7k
Commands
4
Failed
0

Trace

  1. I’ll extract the slide text and speaker notes from the presentation into `show.txt` in the current directory.

  2. rg --files -g 'show.pptx' -g 'AGENTS.md' -g 'show.txt'
    output
    show.pptx
  3. python3 - <<'PY' import importlib.util for m in ['pptx','lxml']: print(m, bool(importlib.util.find_spec(m))) PY
    output
    pptx False
lxml False
  4. unzip -l show.pptx
    output
    Archive:  show.pptx
  Length      Date    Time    Name
---------  ---------- -----   ----
     1355  01-01-1980 00:00   docProps/app.xml
      846  01-01-1980 00:00   docProps/core.xml
     4794  01-01-1980 00:00   docProps/thumbnail.jpeg
     7378  01-01-1980 00:00   ppt/notesMasters/notesMaster1.xml
      292  01-01-1980 00:00   ppt/notesMasters/_rels/notesMaster1.xml.rels
     1568  01-01-1980 00:00   ppt/notesSlides/notesSlide1.xml
     1793  01-01-1980 00:00   ppt/notesSlides/notesSlide2.xml
      447  01-01-1980 00:00   ppt/notesSlides/_rels/notesSlide1.xml.rels
      447  01-01-1980 00:00   ppt/notesSlides/_rels/notesSlide2.xml.rels
     3699  01-01-1980 00:00   ppt/presentation.xml
      816  01-01-1980 00:00   ppt/presProps.xml
     4094  01-01-1980 00:00   ppt/slideLayouts/slideLayout1.xml
     2706  01-01-1980 00:00   ppt/slideLayouts/slideLayout10.xml
     2930  01-01-1980 00:00   ppt/slideLayouts/slideLayout11.xml
     2651  01-01-1980 00:00   ppt/slideLayouts/slideLayout2.xml
     4206  01-01-1980 00:00   ppt/slideLayouts/slideLayout3.xml
     4304  01-01-1980 00:00   ppt/slideLayouts/slideLayout4.xml
     6789  01-01-1980 00:00   ppt/slideLayouts/slideLayout5.xml
     2011  01-01-1980 00:00   ppt/slideLayouts/slideLayout6.xml
     1714  01-01-1980 00:00   ppt/slideLayouts/slideLayout7.xml
     4474  01-01-1980 00:00   ppt/slideLayouts/slideLayout8.xml
     4422  01-01-1980 00:00   ppt/slideLayouts/slideLayout9.xml
      311  01-01-1980 00:00   ppt/slideLayouts/
… 1579 more characters
  5. python3 - <<'PY' import zipfile, xml.etree.ElementTree as ET, re from pathlib import Path p=Path('show.pptx') ns={'a':'http://schemas.openxmlformats.org/drawingml/2006/main','p':'http://schemas.openxmlformats.org/presentationml/2006/main'} with zipfile.ZipFile(p) as z: slides=sorted((n for n in z.namelist() if re.fullmatch(r'ppt/slides/slide\\d+\\.xml',n)), key=lambda n:int(re.search(r'(\\d+)',n).group(1))) notes={int(re.search(r'notesSlide(\\d+)',n).group(1)):n for n in z.namelist() if re.fullmatch(r'ppt/notesSlides/notesSlide\\d+\\.xml',n)} out=[] for i,s in enumerate(slides,1): out.append(f'Slide {i}') root=ET.fromstring(z.read(s)) for sp in root.findall('.//p:sp',ns): texts=[t.text or '' for t in sp.findall('.//a:t',ns)] if texts: out.append(''.join(texts)) n=notes.get(i) if n: out.append('Speaker notes:') nr=ET.fromstring(z.read(n)) for sp in nr.findall('.//p:sp',ns): ph=sp.find('./p:nvSpPr/p:nvPr/p:ph',ns) if ph is not None and ph.attrib.get('type') in ('slidenum','sldImg'): continue texts=[t.text or '' for t in sp.findall('.//a:t',ns)] if texts: out.append(''.join(texts)) out.append('') Path('show.txt').write_text('\\n'.join(out),encoding='utf-8') PY cat show.txt
    output
    Slide 1
Title of the first slide
Subtitle of the first slideThis bit is in italic green
Speaker notes:
I am the notes of the first slide
1

Slide 2
This is the second slide
It has bullet points on itThey’re fun, aren’t they?Especially in a different font like Arial Black at 16 point!
Speaker notes:
These are the notes of the 2nd slideTHIS LINE IS BOLD
2

  6. Extracted the slide text and speaker notes into [show.txt](show.txt).

Output

show.txt 360 B
Slide 1
Title of the first slide
Subtitle of the first slideThis bit is in italic green
Speaker notes:
I am the notes of the first slide
1

Slide 2
This is the second slide
It has bullet points on itThey’re fun, aren’t they?Especially in a different font like Arial Black at 16 point!
Speaker notes:
These are the notes of the 2nd slideTHIS LINE IS BOLD
2

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

Time
15s
New tokens
15.0k
Total input
90.5k
Output tokens
0.4k
Commands
5
Failed
0

Trace

  1. I’ll inspect the available file tools and the presentation, then write the extracted slide text and speaker notes to `show.txt` in this directory.

  2. rg --files -g 'show.pptx' -g 'AGENTS.md'
    output
    show.pptx

  3. I’m using the `deft` workflow for this PowerPoint extraction; I’ll check its instructions for preserving notes and then create the text file.

  4. 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
  5. deft slides
    output
    deft slides - commands (input -> output)
  info     pptx -> report  Slide count, titles, images, notes
  from-md  md -> pptx      Make a deck: each heading or --- starts a slide [beta]
  to-md    pptx -> md      Outline in Markdown: a heading per slide, bullets, images
  to-pdf   pptx -> pdf     Handout PDF of slide content [beta]
  to-text  pptx -> txt     Extract the text of every slide, form feed between slides
More: deft slides <command> --help
  6. deft slides to-text --help
    output
    deft slides to-text - Extract the text of every slide, form feed between slides
Same as: deft convert

Usage: deft slides to-text <pptx>... [-o <txt|dir>] [options]
In: pptx (each handled separately)   Out: txt (default <name>.<ext>)

Options:
  --notes                      Include speaker notes
Common: -o --force --dry-run --json -j   (deft help common)

Examples:
  deft slides to-text deck.pptx --notes -o deck.txt

Result fields: slides
  7. deft slides to-text show.pptx --notes -o show.txt
    output
    ok slides.to-text show.pptx -> show.txt  2 slides  0.00s

  8. Extracted the text from both slides, including speaker notes, into [show.txt](show.txt).

Output

show.txt 326 B
Title of the first slide

Subtitle of the first slide

This bit is in italic green


I am the notes of the first slide



This is the second slide

It has bullet points on it

They’re fun, aren’t they?

Especially in a different font like Arial Black at 16 point!


These are the notes of the 2nd slide
THIS LINE IS BOLD

Claude Opus 5.5Without deftPassed2026-11-07

Time
16s
New tokens
10.9k
Total input
43.1k
Output tokens
1.5k
Commands
0
Failed
0

Trace

    Output

    show.txt 398 B
    ===== Slide 1 =====
Title of the first slide
Subtitle of the first slide
This bit is in italic green
--- Speaker Notes ---
I am the notes of the first slide

===== Slide 2 =====
This is the second slide
It has bullet points on it
They’re fun, aren’t they?
Especially in a different font like Arial Black at 16 point!
--- Speaker Notes ---
These are the notes of the 2nd slide
THIS LINE IS BOLD

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

    Time
    10s
    New tokens
    10.0k
    Total input
    59.4k
    Output tokens
    0.5k
    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
      show.txt 326 B
      Title of the first slide

Subtitle of the first slide

This bit is in italic green


I am the notes of the first slide



This is the second slide

It has bullet points on it

They’re fun, aren’t they?

Especially in a different font like Arial Black at 16 point!


These are the notes of the 2nd slide
THIS LINE IS BOLD

      Claude CodeWithout deftNot run yet

      Claude CodeWith deftNot run yet