social-carousel-design

Warn

Audited by Gen Agent Trust Hub on Sep 13, 2026

Risk Level: MEDIUMDATA_EXFILTRATIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [DATA_EXFILTRATION]: The file references/behance-social-reference-extraction.md contains a hardcoded local file path /Users/akhilaston/Documents/Ai_Hub/, which exposes the author's local username and directory structure.
  • [DYNAMIC_EXECUTION]: The skill is designed to generate and execute Python scripts (e.g., generate_post.py) and HTML/JS code (rendered via Playwright) based on external references. This involves writing source code to disk and executing it at runtime.
  • [COMMAND_EXECUTION]: The instructions involve executing system commands via ffmpeg for video transcoding and running generated scripts through a terminal-runnable workflow.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from external sources including Behance project pages, brand websites, and LinkedIn content calendars.
  • Ingestion points: External URLs (Behance, brand sites) and social media calendar text.
  • Boundary markers: None explicitly defined in the provided instructions.
  • Capability inventory: File system writes, script execution (Python), browser automation (Playwright), and system command execution (ffmpeg).
  • Sanitization: There are no explicit instructions for sanitizing or escaping the content extracted from external sources before it is used in code generation or rendering.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Sep 13, 2026, 07:35 PM
Security Audit — agent-trust-hub — social-carousel-design