kernel-security
Pass
Audited by Gen Agent Trust Hub on Jun 27, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides educational content and functional examples for Linux security modules (LSMs) like SELinux and AppArmor, as well as hardening techniques like Intel CET and ARM PAC. All provided commands use standard system utilities for their intended security purposes.\n- [INDIRECT_PROMPT_INJECTION]: The skill defines workflows for processing external data, which represents a potential attack surface for indirect prompt injection.\n
- Ingestion points: The skill analyzes audit logs, process output from strace, and external CVE information from the NVD and distro security trackers.\n
- Boundary markers: There are no specific delimiters or boundary markers defined in the instructions for handling these external data sources.\n
- Capability inventory: The skill possesses the capability to compile code via gcc and make, and to write security policy files to the system.\n
- Sanitization: No explicit sanitization or validation logic is provided for the ingestion of external data.
Audit Metadata