The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it is designed to ingest and process data from external files.
Ingestion points: Data is loaded from external files using read_xlsx, load_workbook, and csv.reader as shown in SKILL.md.
Boundary markers: The provided code templates and tool definitions do not include boundary markers or specific instructions to ignore malicious prompts embedded within spreadsheet cells.
Capability inventory: The skill possesses significant capabilities including file system write access and the ability to execute generated Python code.
Sanitization: There is no evidence of content sanitization or validation to filter out executable instructions from the input data.
[COMMAND_EXECUTION]: The skill operates by generating and executing Python code at runtime to perform spreadsheet tasks.
Evidence: The overview section explicitly states that the agent will generate and execute openpyxl code based on user prompts.
[EXTERNAL_DOWNLOADS]: The documentation recommends the installation of the openpyxl library.
Evidence: The installation section references pip install openpyxl. This is a well-known and standard library for Excel processing.

xlsx-manipulation