alignment

Pass

Audited by Gen Agent Trust Hub on Nov 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes content from external web pages, which are untrusted data sources.
  • Ingestion points: The script measure.js reads text and element positions from the browser's Document Object Model (DOM).
  • Capability inventory: The skill writes a local report to the .alignment/ directory and executes JavaScript within the browser context via page.evaluate or similar.
  • Boundary markers: Instructions mandate that the agent writes structured data only to report.js and avoids including custom markup in the final report.
  • Sanitization: Text content extracted from analyzed web pages is not explicitly sanitized before being written into the report data file.
  • [DYNAMIC_EXECUTION]: The skill creates a report.js file at runtime to supply measured data to the visual report template.
  • The file is generated based on a predefined template and contains the results of the layout audit.
  • [EXTERNAL_DOWNLOADS]: The report.html template references typography assets from Google's public font services.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 4, 2026, 08:23 AM