Skills
skills/7spade/black-tortoise/docx/Gen Agent Trust Hub

docx

Fail

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: HIGHCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • COMMAND_EXECUTION (HIGH): The script ooxml/scripts/pack.py uses subprocess.run to call soffice (LibreOffice). This creates a risk of command exploitation or interaction with a complex external tool using agent-controlled files.
  • PROMPT_INJECTION (HIGH): The skill possesses a significant Indirect Prompt Injection surface (Category 8) as its core function is to process external documents.
  • Ingestion points: ooxml/scripts/unpack.py extracts and reads content from user-provided .docx, .pptx, and .xlsx files.
  • Boundary markers: None detected. The skill does not use delimiters or instructions to ignore instructions embedded within the document XML.
  • Capability inventory: Broad file system access via extractall and external command execution via subprocess.run in pack.py.
  • Sanitization: The extractall method in ooxml/scripts/unpack.py is vulnerable to Zip Slip, which can overwrite files outside the target directory. Additionally, lxml.etree in ooxml/scripts/validation/docx.py is used without explicit protections against XXE (XML External Entity) attacks.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Feb 16, 2026, 06:50 AM