| name | pdf2img |
| description | Convert a PDF file into per-page PNG images and save them into a dedicated subfolder inside the artifacts directory. Use this skill whenever the user provides a PDF path (e.g. ~/Downloads/6660-SPS.pdf) and asks to extract, convert, or save its pages as images. The output folder is named after the PDF file stem (e.g. artifacts/6660-SPS/).
|
pdf2img
Convert every page of a PDF into individual PNG images.
Workflow
-
Derive the output directory from the PDF file stem:
pdf_path = ~/Downloads/6660-SPS.pdf
stem = 6660-SPS
output_dir = <ARTIFACTS_DIR>/6660-SPS/
-
Run the bundled script via execute_code or bash:
python skills/pdf2img/scripts/pdf2img.py "<pdf_path>" "<output_dir>"
The script auto-installs pymupdf if it is missing.
-
Report results – list the saved image paths and total page count to the user.
Script
scripts/pdf2img.py — accepts two positional arguments:
| Argument | Description |
|---|
pdf_path | Path to the source PDF (supports ~) |
output_dir | Destination folder (created if absent) |
Default resolution is 150 DPI (clear enough for most documents).
Images are named page_001.png, page_002.png, … in order.
Dependencies
pymupdf (fitz) — auto-installed by the script if missing.
Example
User says: "~/Downloads/6660-SPS.pdf 파일을 이미지로 변환해줘"
import subprocess, sys
script = "skills/pdf2img/scripts/pdf2img.py"
pdf = os.path.expanduser("~/Downloads/6660-SPS.pdf")
outdir = os.path.join(ARTIFACTS_DIR, "6660-SPS")
subprocess.run([sys.executable, script, pdf, outdir], check=True)
