skills/l-yifan/skills/pptx/Gen Agent Trust Hub

pptx

Warn

Audited by Gen Agent Trust Hub on Feb 24, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The script scripts/office/soffice.py performs runtime compilation and library injection. It dynamically writes a C source file (lo_socket_shim.c) to the temporary directory, compiles it into a shared object using gcc, and then uses the LD_PRELOAD environment variable to inject this library into the soffice (LibreOffice) process. This is designed to bypass Unix socket restrictions in restricted execution environments.- [COMMAND_EXECUTION]: The skill executes various system utilities through the subprocess module. Specifically, scripts/office/soffice.py and scripts/thumbnail.py invoke gcc, soffice, and pdftoppm to process presentation files and manage the socket shim.- [PROMPT_INJECTION]: The skill contains an indirect prompt injection surface documented in SKILL.md and scripts/thumbnail.py. It processes untrusted .pptx files and transmits extracted text or rendered slide images to subagents for visual QA. The workflow lacks explicit boundary markers or content sanitization to prevent subagents from following malicious instructions embedded within the user-provided presentation data.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 24, 2026, 08:16 AM