pymupdf

Render PDF pages to images, extract embedded images, annotate PDFs, and perform advanced PDF inspection using pymupdf (fitz). Use for tasks such as exporting...

Security Scan

The skill is internally consistent: its code, CLI, and runtime instructions match the stated purpose of rendering and extracting images from PDFs and do not request unrelated credentials or system access.

Purpose & Capability

Name, description, SKILL.md, and the included scripts all align: the CLI opens a local PDF, reports info, renders pages to images, and extracts embedded image streams. The only required binary is python3 and no credentials or unrelated tools are requested.

Instruction Scope

SKILL.md instructs running the provided CLI and to ensure the PDF path is local. It does not direct the agent to read unrelated files, access environment variables, or transmit data externally. It does recommend installing the dependency with pip if missing, which is expected for a Python-based tool.

Install Mechanism

There is no automated install spec (instruction-only), so nothing is written to disk by the skill itself. The README suggests 'pip install pymupdf' when needed — a normal, low-risk developer action. No downloads from arbitrary URLs or archive extraction are present.

Credentials

No environment variables, credentials, or config paths are required. The script only accesses the input PDF path and writes to the user-specified output directory, which is proportional to the declared functionality.

Persistence & Privilege

The skill does not request always:true and makes no modifications to other skills or global agent configuration. Normal autonomous invocation is allowed by default but this is standard and not combined with other privileges.

Assessment

This skill appears coherent and limited to operating on local PDF files. Before installing or running it, consider: 1) The script will open and process whatever PDF path you provide and will write files to the specified output directory — verify paths and permissions. 2) The SKILL.md recommends installing the PyMuPDF package from PyPI; install packages in a virtualenv/container and pin versions to reduce supply-chain risk. 3) Parsing untrusted PDFs can trigger vulnerabilities in native libraries; avoid running on untrusted documents on sensitive systems. 4) There are no network calls or credential requests in the code, so it does not exfiltrate data by itself, but always review any third-party package you install. If you want extra safety, run the tool in an isolated environment (VM, container, or dedicated sandbox) and inspect the PyMuPDF package version before installation.

Use scripts/pymupdf_cli.py for deterministic pymupdf operations instead of ad-hoc fitz snippets.

Workflow

1. Confirm the PDF file path is accessible locally.
2. Run the desired command (export-images, extract-images, info).
3. Inspect the output directory or stdout.

Command Guide

• Export pages as images:
  • python scripts/pymupdf_cli.py export-images --input <file.pdf> --output-dir <dir/>
  • Override format: --format png (default), jpg, or ppm.
  • Override resolution: --dpi 300 (default: 150).
  • Restrict to specific pages: --pages 0 1 2.
• Extract images embedded inside a PDF:
  • python scripts/pymupdf_cli.py extract-images --input <file.pdf> --output-dir <dir/>
  • Optionally restrict to specific pages: --pages 0 1 2.
• Inspect page dimensions and basic document info:
  • python scripts/pymupdf_cli.py info --input <file.pdf>

Operational Rules

• Pages are always 0-indexed in all commands.
• export-images renders each page as a raster image at the specified DPI.
• extract-images saves raw image streams embedded in the PDF (e.g. photos, logos); output filenames include page index and image index.
• Install dependency if missing: pip install pymupdf.
• For text extraction, splitting, merging, or rotating PDFs, use the pypdf skill instead.