PDF to Word Converter
PassAudited by ClawScan on May 1, 2026.
Overview
The skill appears to be a disclosed local PDF conversion wrapper, with the main things to notice being its external ComPDF dependency/downloads and optional handling of PDF passwords.
Before installing, be comfortable with using the ComPDFKitConversion package and first-run downloads from compdf.com, including the large AI model. Run it only on documents you are willing to process with this local SDK, confirm output paths to avoid overwrites, and avoid exposing protected-PDF passwords in shared logs or shells.
Findings (3)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
First use may download and cache provider files, including a large AI model, and the conversion workflow depends on those external assets.
The skill depends on an external Python SDK and first-run downloads of provider-hosted license/model files; this is disclosed and purpose-aligned, but users are relying on external provenance and availability.
pip install ComPDFKitConversion ... script automatically downloads `license.xml` ... automatically download it from: https://download.compdf.com/skills/model/documentai.model
Install the SDK from a trusted source, consider pinning versions or verifying downloads where possible, and allow the first-run network download only if the ComPDF provider dependency is acceptable.
Incorrect or unexpected arguments could write output to the wrong location or cause the local conversion command to behave differently than intended.
The skill is intended to run a local command using user-provided format, input path, output path, and options. That is expected for a converter, but the user should verify arguments and output destinations.
python "${CLAUDE_SKILL_DIR}/scripts/pdf-to-word-docx.py" $ARGUMENTSReview the input file, output path, and conversion options before running, especially when paths contain spaces or special characters.
A PDF password may be visible in local shell history, process listings, or agent logs depending on the environment.
The script can accept a password for protected PDFs. This is purpose-aligned and no logging or transmission is evidenced, but it is still sensitive input.
parser.add_argument("--password", default="", help="PDF password")Use the password option only in a trusted local environment and avoid sharing logs or command transcripts that contain document passwords.
