frontend-slides

Pass

Audited by Gen Agent Trust Hub on Mar 12, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted data from users or external files to generate slides.
  • Ingestion points: 'Discover Content' step and 'PPT / PPTX Conversion' section in SKILL.md.
  • Boundary markers: None specified.
  • Capability inventory: Writing HTML files (Step 4), executing system commands to open them (Step 7), and potential use of browser automation for validation (Step 6).
  • Sanitization: None mentioned.
  • [COMMAND_EXECUTION]: The skill uses shell commands (open, xdg-open, start) to launch the generated presentation in a browser, which is an intended part of its workflow.
  • [EXTERNAL_DOWNLOADS]: The skill identifies the 'python-pptx' library as a dependency for PowerPoint conversion and may suggest its installation from the standard Python Package Index (PyPI).
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 12, 2026, 08:40 AM