vida-perspective

Pass

Audited by Gen Agent Trust Hub on Oct 26, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is designed for character emulation and analysis based on a provided local dataset. It contains no executable code, network requests, or attempts to access sensitive system information.- [PROMPT_INJECTION]: The skill uses roleplay instructions to adopt a specific persona. These instructions are contained within a clear character framework and are accompanied by explicit prohibitions against generating illegal or harmful content.- [INDIRECT_PROMPT_INJECTION]: The skill processes a large volume of local text data to inform its responses. While this creates a theoretical surface for data-driven influence, the risk is mitigated by the skill's lack of external capabilities (like file system writes or network access) and the use of locally-stored, static archives.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 26, 2026, 09:32 AM