The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/chart_handler.py executes shell commands to render Mermaid diagrams.
Evidence: It calls subprocess.run() to execute the mmdc (Mermaid CLI) command.
Safety Context: The command is constructed using a list of arguments rather than a raw shell string, and shell=True is not used, which follows security best practices to prevent command injection. The execution is gated by the presence of a local Mermaid installation.
[EXTERNAL_DOWNLOADS]: The main script scripts/md2word.py downloads images from external URLs provided in the Markdown source.
Evidence: The download_external_image function uses urllib.request.urlopen to fetch remote image data.
Safety Context: This is a documented primary feature (v0.5.0) intended to allow users to embed remote images into their Word documents. It includes a standard User-Agent and a 20-second timeout to prevent hanging or basic request blocking.
[SAFE]: The skill uses yaml.safe_load in scripts/config.py for parsing configuration files, which prevents arbitrary code execution during YAML deserialization.
[SAFE]: Recommended dependencies (python-docx, Pillow, beautifulsoup4, PyYAML, and @mermaid-js/mermaid-cli) are well-known, standard libraries from official registries (PyPI and NPM).

md2word