The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructions involve executing several shell commands via the Bash tool, including the use of google-chrome --headless for PDF generation and libreoffice --headless for document conversion. It also executes local Python scripts to generate visual elements.
[EXTERNAL_DOWNLOADS]: The skill requires the installation and use of the python-pptx library and assumes the presence of external binaries such as Google Chrome and LibreOffice on the host system.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it incorporates untrusted user data into HTML templates and command-line arguments.
Ingestion points: Research data and figure descriptions provided by the user are written to poster.html and passed to scripts/generate_schematic.py via shell arguments.
Boundary markers: No boundary markers or 'ignore' instructions are used to separate user data from the skill's operational logic.
Capability inventory: The skill utilizes Bash, Read, and Write tools to perform file system operations and execute system-level applications.
Sanitization: There is no evidence of sanitization or escaping of user-provided text before it is used in shell commands or HTML generation.
[NO_CODE]: The skill references a core functional script, scripts/generate_schematic.py, which is not included in the provided file set, preventing a full security audit of the image generation logic.

pptx-posters