The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill recommends installation via npx openskills from a specific GitHub repository (github.com/okwinds/miscellany). It also depends on standard Python data science libraries such as pandas and openpyxl for Excel processing.
[COMMAND_EXECUTION]: The recalc.py script invokes the soffice (LibreOffice) binary using subprocess.run to perform headless formula recalculations. This execution is gated by the script's specific purpose.
[DYNAMIC_EXECUTION]: In recalc.py, the skill dynamically generates a LibreOffice Basic macro (Module1.xba) and writes it to the filesystem. This macro is then executed via LibreOffice to automate the recalculation and saving of workbooks. While the script uses an isolated profile by default, it can optionally modify the user's permanent LibreOffice configuration.
[INDIRECT_PROMPT_INJECTION]:
Ingestion points: The skill processes external data from .xlsx files through pandas.read_excel, openpyxl.load_workbook, and LibreOffice's calculation engine.
Boundary markers: No explicit boundary markers or instructions to ignore embedded cell content are used during ingestion.
Capability inventory: The skill possesses the ability to execute system commands (soffice) and perform file system operations (writing macros and saving workbooks).
Sanitization: There is no evidence of sanitization or validation of the data read from Excel cells before it is processed by the agent.

xlsx-offline