html-to-ppt-marp

Pass

Audited by Gen Agent Trust Hub on Mar 31, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The Python code provided in SKILL.md uses subprocess.run to execute the marp CLI tool. While it uses a list of arguments to mitigate shell injection, it establishes a pattern where the agent executes system commands based on input data.\n- [PROMPT_INJECTION]: The skill acts as a surface for indirect prompt injection by ingesting untrusted data (such as slide titles, points, and Markdown content) and interpolating it into templates without sanitization.\n
  • Ingestion points: External data enters the agent context via the create_presentation and markdown_to_pptx function parameters in SKILL.md.\n
  • Boundary markers: No delimiters or 'ignore embedded instructions' markers are used when interpolating user data into the Markdown structure.\n
  • Capability inventory: The skill scripts utilize tempfile for writing data to disk and subprocess.run to execute external tools on that data.\n
  • Sanitization: There is no evidence of escaping or validation performed on the input strings before they are combined into the Markdown file.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 31, 2026, 02:03 AM
Security Audit — agent-trust-hub — html-to-ppt-marp