The Agent Skills Directory

[PROMPT_INJECTION]: Indirect prompt injection surface detected because the skill processes untrusted external data.
Ingestion points: The skill reads data from external CSV files (data_csv) and document templates (template_path) via the mail_merge and fill_template functions.
Boundary markers: No explicit delimiters or instructions to the model to ignore embedded commands are present in the template rendering process.
Capability inventory: The skill possesses file system read and write capabilities for various document formats through docxtpl, openpyxl, and the csv module.
Sanitization: The skill does not perform input validation or sanitization on external data before it is interpolated into templates using Jinja2 logic.

template-engine