skills/read2017/cabinet/life-cabinet/Gen Agent Trust Hub

life-cabinet

Pass

Audited by Gen Agent Trust Hub on Jun 20, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is designed to facilitate complex decision-making through simulated advisory sessions and structured reporting. It relies on local file storage for session persistence and user profiling.
  • [DATA_EXPOSURE]: The instructions involve collecting personal and financial data to improve advice quality. However, the skill explicitly warns against collecting high-sensitivity information such as passwords or government IDs, and specifies that data should be stored locally in the outputs/life-cabinet/ directory. It maintains a user profile to persist data between sessions but requires user confirmation before updates.
  • [PROMPT_INJECTION]: When generating HTML reports that include user-supplied content, the skill contains explicit instructions to escape special characters (<, >, &). This practice mitigates the risk of indirect prompt injection or cross-site scripting (XSS) within the rendered report files. It also uses 'context freezing' to define a clear boundary between gathered facts and the subsequent persona analysis.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 20, 2026, 07:22 AM
Security Audit — agent-trust-hub — life-cabinet