meta-agent
Pass
Audited by Gen Agent Trust Hub on Feb 23, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [NO_CODE] (SAFE): The skill is entirely composed of Markdown text and metadata. It does not contain scripts (.py, .js, .sh), binary files, or executable instructions.\n- [EXTERNAL_DOWNLOADS] (SAFE): No remote resources, external scripts, or package installations are present. All references to other skills use local relative paths.\n- [DATA_EXFILTRATION] (SAFE): There are no commands for network activity or accessing sensitive system files such as credentials or SSH keys.\n- [PROMPT_INJECTION] (SAFE): The instructional content focuses on high-level strategy and does not contain directives to ignore system constraints or bypass safety filters.
Audit Metadata