aris-paper-slides

Warn

Audited by Gen Agent Trust Hub on May 6, 2026

Risk Level: MEDIUMCREDENTIALS_UNSAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [CREDENTIALS_UNSAFE]: The skill checks for the existence of ~/.claude/feishu.json to enable optional notifications. Accessing hidden application-specific credential files in the home directory is a security risk.
  • [COMMAND_EXECUTION]: The skill dynamically generates a Python script (generate_pptx.py) and executes it using python3 to convert LaTeX frames to PowerPoint slides.
  • [EXTERNAL_DOWNLOADS]: The skill executes pip install python-pptx to download and install the python-pptx library if it is not present in the environment.
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it reads and processes untrusted user-provided LaTeX files (paper/sections/*.tex) and uses this content to build slide outlines and speaker notes.
  • Ingestion points: paper/sections/*.tex and paper/main.tex.
  • Boundary markers: Absent; no specific instructions to ignore embedded commands in the source paper.
  • Capability inventory: Shell access for compilation and package management (bash), file writing (Write), and agent-based reasoning (mcp__codex__codex).
  • Sanitization: Absent; no validation or escaping of LaTeX content before processing.
Audit Metadata
Risk Level
MEDIUM
Analyzed
May 6, 2026, 01:42 AM
Security Audit — agent-trust-hub — aris-paper-slides