The Agent Skills Directory

[COMMAND_EXECUTION]: The recalc.py script executes system commands via subprocess.run to interact with LibreOffice and manage execution timeouts. The script constructs command lists that include the soffice binary and system timeout utilities (timeout or gtimeout) to process the target Excel files.
[DYNAMIC_EXECUTION]: The skill dynamically generates a LibreOffice Basic macro and saves it to the local filesystem to facilitate spreadsheet recalculation. The setup_libreoffice_macro function in recalc.py writes XML-formatted Basic code to a file named Module1.xba within the user's LibreOffice configuration directory and subsequently triggers its execution.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external spreadsheet files (Excel, CSV, TSV), which creates a potential attack surface for instructions embedded in data processed by the agent.
Ingestion points: The skill uses pd.read_excel() and load_workbook() to read user-supplied or external files.
Boundary markers: No explicit boundary markers or warnings to ignore embedded instructions are provided in the skill guidelines.
Capability inventory: The skill has capabilities including local shell command execution (via recalc.py), filesystem access, and the ability to execute arbitrary Python code for data manipulation.
Sanitization: There is no evidence of sanitization or validation of cell content before it is processed or used in calculations.

xlsx