pptx-author

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • Local Script Execution: The skill generates and executes Python scripts using the python-pptx library. This is a standard pattern for document generation but involves dynamic execution of agent-created code.
  • File System Access: The skill interacts with the local file system by reading templates and writing output to a designated ./out/ directory. It also sources data from local spreadsheets.
  • Indirect Prompt Injection Surface: The skill processes data from external sources (such as ./out/model.xlsx) to generate slide content. As with any data ingestion, this presents a surface for instructions embedded in data to be processed, though the skill includes explicit instructions against external data exfiltration.
  • Ingestion points: Data is pulled from ./out/model.xlsx (SKILL.md).
  • Boundary markers: No specific delimiters or safety warnings for embedded content are defined.
  • Capability inventory: Executes Python via Bash, writes .pptx files to disk (SKILL.md).
  • Sanitization: No explicit sanitization or validation of the input data is mentioned.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 12:33 AM
Security Audit — agent-trust-hub — pptx-author