retro-analysis
Pass
Audited by Gen Agent Trust Hub on Mar 3, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [NO_CODE]: This skill consists entirely of natural language instructions and does not include any executable scripts or binary code.\n- [SAFE]: No evidence of malicious behavior, credential theft, or persistence mechanisms was found.\n- [PROMPT_INJECTION]: The instructions are focused on sprint data analysis and do not contain patterns typical of prompt injection or system instruction bypass.\n- [DATA_EXFILTRATION]: The skill is designed to process sprint-related metrics from Jira and Miro as part of its primary function; no mechanisms for unauthorized data exfiltration or sensitive file access were identified.
Audit Metadata