research-paper-pdf-translator

This skill appears to do what it says, but consider these practical safety points before installing: - Data leakage: SKILL.py prints the entire extracted paper text and the absolute PDF path to stdout and asks the agent to translate it. If your agent sends prompts or content to a remote LLM or logs stdout, the full paper and filesystem path could be transmitted or recorded. Avoid using with confidential PDFs unless you control where model invocations and logs go. - Temporary file handling: The script writes a fixed file name (temp_paper_for_translation.txt) in the same directory as the PDF. That can overwrite files with the same name or leave artifacts if the script crashes. Prefer running in a copy of the directory or modify the script to use a secure temp file (e.g., tempfile in /tmp) or a unique name. - Write permission: The skill asks the agent to save the resulting markdown in the same directory; ensure you trust the agent's write permissions for that location. - Dependency: You must install poppler-utils (pdftotext) for reliable conversion. If you need stronger privacy guarantees, run the script in an isolated environment, modify it to use secure temporary files and avoid printing absolute paths, and ensure model invocations remain local or to a trusted endpoint.