pptx-extract: PowerPoint text + image extractor

What it does

Pulls every slide’s text and every embedded image out of a .pptx deck. A .pptx is just a zip archive, so the bundled scripts/extract_pptx.py unzips it in memory, reads the <a:t> text runs from each ppt/slides/slideN.xml (in numeric slide order), maps each slide’s images via its _rels file, copies the referenced media into an output dir, and prints a per-slide report of text plus the absolute paths of the copied images. Speaker notes are included when present.

Usage

python3 ~/.claude/skills/pptx-extract/scripts/extract_pptx.py "/path/to/deck.pptx"

Pass --out <dir> to choose the output directory; otherwise a fresh temp dir under the system temp is created and its path is printed at the top of the report.

After running

The script gives you the text and the image file locations, not the visual analysis. To describe the actual UI/mockups, Read the printed media image paths (the Media: ... lines) so you can see and describe each slide’s images.

Notes

• Standard library only (zipfile, xml.etree.ElementTree, shutil, argparse, tempfile, re). No python-pptx, no pip installs required.
• A .pptx is a zip file; slides are enumerated numerically (slide2 before slide10), not lexically.
• Robust to decks with no media, no text, or missing _rels; exits non-zero with a clear message only if the file is missing or not a valid zip.

Triggers on: “extract pptx”, “read this powerpoint”, “read the .pptx”, “extract slides”, “what’s in this deck”, “pptx”, “/pptx-extract”.