The Agent Skills Directory

[COMMAND_EXECUTION]: The skill operates by generating and executing Python scripts (utilizing openpyxl or the included xlsx_tool.py) to produce binary Excel files. This is the primary intended functionality for creating editable spreadsheets.
[INDIRECT_PROMPT_INJECTION]: The skill processes external user input, such as financial assumptions and model requirements, to generate workbook content. While this presents a theoretical surface for injection, the provided helper script includes appropriate XML escaping to prevent data-level exploits.
Ingestion points: SKILL.md defines requirements for user-provided data including assumptions, outputs, and model structure.
Boundary markers: The skill does not explicitly define delimiters for untrusted data in its processing logic.
Capability inventory: The skill involves file system writes and Python code execution within the agent's environment.
Sanitization: scripts/xlsx_tool.py utilizes xml.sax.saxutils.escape to sanitize cell values and sheet names, effectively mitigating XML-based injection risks in the resulting Excel files.

excel-model