maintaining-engagement-state
Pass
Audited by Gen Agent Trust Hub on Aug 7, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill is purely instructional and focuses on professional best practices for maintaining engagement state and ensuring provenance in security reporting.
- [NO_CODE]: The skill contains only documentation templates and guidelines; it does not include any scripts, executable commands, or tool definitions.
- [PROMPT_INJECTION]: There are no patterns suggesting attempts to override model instructions, bypass safety filters, or extract system prompts.
- [DATA_EXPOSURE]: While the skill discusses recording credentials and access, it does so through hypothetical documentation examples. It does not contain hardcoded secrets or instructions to access sensitive files on the execution environment.
Audit Metadata