Skill detail
pptx-author
Direct headless PPTX authoring for deliverable slide decks.
Inspect before use
Automated review checks relevance, not safety or endorsement. Read the source instructions before using this skill.
SKILL.md
The saved excerpt is a snapshot from review. The external source remains the complete and most current version.
---
name: pptx-author
description: Build PowerPoint decks headless with python-pptx.
version: 1.0.0
author: Anthropic (adapted by Nous Research)
license: Apache-2.0
platforms: [linux, macos, windows]
metadata:
hermes:
tags: [powerpoint, pptx, python-pptx, presentation, finance]
related_skills: [excel-author, powerpoint]
---
# pptx-author
Produce a .pptx file on disk using `python-pptx`. Use when you need to deliver a deck as a file artifact, not drive a live PowerPoint session.
Adapted from Anthropic's `pptx-author` and `pitch-deck` skills in [anthropics/financial-services](https://github.com/anthropics/financial-services). The MCP / Office-JS branches of the originals are dropped — this assumes headless Python.
For the broader, already-shipped PowerPoint authoring skill (slides, speaker notes, embeds, media), see the built-in `powerpoint` skill. This skill is a lighter-weight pattern tuned for model-backed decks (pitch decks, IC memos, earnings notes) where every number must trace to a source workbook.
## Output contract
- Write to `./out/<name>.pptx`. Create `./out/` if it does not exist.
- Return the relative path in your final message.
## Setup
```bash
pip install "python-pptx>=0.6"
```
## Core conventions
### One idea per slide
Title states the takeaway; body supports it. A slide titled "Q3 Revenue" is weak; "Revenue growth accelerated to 14% Y/Y in Q3" is strong.
### Every number traces to the model
If a figure on a slide came from `./out/model.xlsx`, footnote the sheet and cell.
```
Revenue: $1,250M (Source: model.xlsx, Inputs!C3)
```
Never transcribe numbers from memory or from a summary — open the workbook, read the named range, and bind the deck value to it programmatically when you can.
### Use the firm template when one is mounted
If `./templates/firm-template.pptx` exists, load it so the deck inherits branded colors, fonts, and master layouts.
```python
from pptx import Presentation
from pathlib import Path
template = Path("./templates/firm-template.pptx")
prs = Presentation(str(template)) if template.exists() else Presentation()
```
### Charts: PNG-from-model beats native pptx charts
When fidelity matters (the model's chart styling must match the deck exactly), render the chart to PNG from the source workbook and embed the image. Native `pptx.chart` charts are fragile and often don't match firm conventions.
```python
from pptx.util import Inches
slide.shapes.add_picture("./out/charts/football_field.png",
Inches(1), Inches(2),
width=Inches(8))
```
### No external sends
This skill writes a file. It never emails, uploads, or posts. Orchestration layers handle delivery.
## Skeleton
```python
from pptx import Presentation
from pptx.util import Inches, Pt
from pptx.dml.color import RGBColor
from pathlib import Path
template = Path("./templates/firm-template.pptx")
prs = Presentation(str(template)) if template.exists() else Presentation()
# Title slide
slide = prs.slides.add_slide(prs.slide_layouts[0])
slide.shapes.title.text = "Project Aurora — Strategic Alternatives"
slide.placeholders[1].text = "Preliminary Discussion Materials"
# Valuation summary slide (title-only layout)
slide = prs.slides.add_slide(prs.slide_layouts[5])
slide.shapes.title.text = "Valuation implies $38–$52 per share across methodologies"
# Add a table bound to model outputs
rows, cols = 5, 4
tbl_shape = slide.shapes.add_table(rows, cols,
Inches(0.5), Inches(1.5),
Inches(9), Inches(3))
tbl = tbl_shape.table
headers = ["Methodology", "Low ($)", "Mid ($)", "High ($)"]
for c, h in enumerate(headers):
tbl.cell(0, c).text = h
# In a real deck, read these from the model workbook with openpyxl
data = [
("Trading comps", "35", "41", "48"),
("Precedent M&A", "39", "45", "52"),
("DCF (base)", "36", "43", "51"),
("LBO (10% IRR)", "33", "38", "44"),
]
for r, row in enRead the full source on GitHub (opens external page)