template-engine

Pass

Audited by Gen Agent Trust Hub on Mar 13, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides instructions to execute package installation and script execution for document generation.
  • Evidence: Instructions include pip install docxtpl openpyxl, pip install weasyprint, and several Python scripts for file processing.
  • Context: These are standard operations for the stated purpose of a template engine and use well-known, legitimate libraries.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external data (CSV, JSON, Spreadsheets) and injects it directly into templates (Jinja2, docxtpl).
  • Ingestion points: SKILL.md (e.g., open(csv_path), json.load(f), csv.DictReader(f)).
  • Boundary markers: None present in the code snippets to separate data from instructions.
  • Capability inventory: File system write access (doc.save(), out.write(), write_pdf()).
  • Sanitization: The guidelines mention sanitizing user-provided data, but the provided code examples do not implement specific sanitization or escaping logic before rendering.
  • Risk: If a CSV or JSON file contains malicious instructions (e.g., markdown injection or HTML-based prompt injection), the generated documents could influence the behavior of the agent reading them or lead to cross-site scripting (XSS) if the HTML output is rendered in a browser.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 13, 2026, 09:16 PM