share

Pass

Audited by Gen Agent Trust Hub on Jul 29, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute a local Python script (scripts/render_carousel.py) via uv run to generate visual assets from report data.\n- [EXTERNAL_DOWNLOADS]: The skill facilitates the installation of the Chromium browser using the Playwright library. Playwright is a standard and well-known browser automation tool from Microsoft.\n- [PROMPT_INJECTION]: The skill ingests external reports to generate platform-specific content, presenting a surface for indirect prompt injection where malicious data in a report could influence the agent's output.\n
  • Ingestion points: Processes report files specified by the user in the SKILL.md workflow.\n
  • Boundary markers: None explicitly defined in the instructions to separate report content from the agent's system prompt.\n
  • Capability inventory: The skill writes PNG and PDF files to the local file system and executes shell commands for rendering.\n
  • Sanitization: The render_carousel.py script implements html.escape() for all user-provided JSON fields, mitigating risks of XSS or injection within the headless browser environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 29, 2026, 12:16 PM
Security Audit — agent-trust-hub — share