pitch-deck

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • Command Execution for Validation: The skill utilizes command-line tools such as soffice (LibreOffice) and pdftoppm to convert PowerPoint presentations into images. While shell execution is a security consideration, it is used here specifically for the purpose of visual validation to ensure formatting standards are met.
  • Indirect Prompt Injection Surface: The workflow involves processing data from various external sources, including Excel, CSV, PDF, and web materials (Ingestion points: SKILL.md). This represents a potential security consideration as it ingests untrusted data. However, the skill directs these capabilities solely toward the structured task of content mapping and template population (Capability inventory: soffice, pdftoppm, and file manipulation via python-pptx in SKILL.md and xml-reference.md).
  • Programmatic Document Modification: The skill provides instructions for modifying PowerPoint files through both direct XML editing and the python-pptx library. This is a security finding related to how the agent handles document structure, but it is provided as a reference to ensure precise formatting and structural integrity in the final output.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 03:06 PM
Security Audit — agent-trust-hub — pitch-deck