scenario-probe
Pass
Audited by Gen Agent Trust Hub on Aug 26, 2026
Risk Level: SAFE
Full Analysis
- [DATA_EXPOSURE_AND_EXFILTRATION]: The skill operates in a read-only capacity regarding the audited artifacts. It contains specific instructions to avoid writing outputs to the audited directory, requiring user-specified paths for any generated reports or trace files.
- [INDIRECT_PROMPT_INJECTION]: The skill has a data ingestion surface as it processes user-provided instruction text. However, this content is processed for analysis and simulation rather than direct execution. The risk of the auditor being influenced by malicious text in the audited file is inherent to the tool's purpose but is mitigated by the structured simulation methodology.
Audit Metadata