sgds-writing

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists entirely of documentation and stylistic guidelines for writing government digital content. No executable code or malicious patterns were detected.
  • [PROMPT_INJECTION]: Heuristic detections of concealment are false positives. The guide recommends removing subjective adjectives (like 'important' or 'easy') from user interface copy to improve clarity and trust; this is a standard plain-language design principle and does not constitute an attempt to hide agent behavior.
  • [EXTERNAL_DOWNLOADS]: The reference materials include instructions for developers to configure spell-checking tools using standard community packages (e.g., '@cspell/dict-en-gb'). These are environment setup instructions for developers and do not pose a runtime risk to the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 04:55 AM