Skills
Navigate
skills/edwinhu/workflows/law-econ-docx/Gen Agent Trust Hub

law-econ-docx

Pass

Audited by Gen Agent Trust Hub on Oct 5, 2026

Risk Level: SAFEDYNAMIC_CONTEXT_INJECTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [DYNAMIC_CONTEXT_INJECTION]: The skill uses the dynamic context syntax (!) in SKILL.md to execute the skill-toc utility at load time. This command generates a directory listing for navigation and does not exfiltrate data or access sensitive files.
  • [COMMAND_EXECUTION]: The scripts build_le_docx.py and make_le_template.py invoke pandoc using subprocess.run with list-formatted arguments. This is a legitimate and secure method for document rendering within the skill's purpose.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided markdown files. The content is processed deterministically by Pandoc and Python scripts rather than being interpolated into an LLM prompt. The process includes assembly steps like frontmatter stripping and namespacing, and the output is a binary DOCX file, mitigating injection risks.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 5, 2026, 10:36 AM