The Agent Skills Directory

[COMMAND_EXECUTION]: The recalc.py script executes system-level binaries using the subprocess module to perform office document operations.
Evidence: The script calls soffice (LibreOffice), timeout, and gtimeout to trigger formula recalculation in a headless environment.
[COMMAND_EXECUTION]: The skill modifies the local application environment by writing configuration files to the user's home directory.
Evidence: The setup_libreoffice_macro function in recalc.py creates a StarBasic macro file (Module1.xba) in the LibreOffice configuration path (~/.config/libreoffice/ on Linux or ~/Library/Application Support/LibreOffice/ on macOS).
[PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by processing untrusted data from external Excel files.
Ingestion points: User-supplied .xlsx files read via pandas and openpyxl as instructed in SKILL.md.
Boundary markers: Absent; there are no instructions to use delimiters or warnings for the agent to ignore instructions embedded in spreadsheet cells.
Capability inventory: The skill has the ability to execute shell commands (via recalc.py) and perform file system operations.
Sanitization: Absent; the skill does not include logic to validate or sanitize content extracted from spreadsheet cells before it enters the agent's context.

xlsx