aegis-linux-baseline
Pass
Audited by Gen Agent Trust Hub on Aug 31, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute a variety of standard Linux diagnostic and administrative commands (e.g.,
awk,grep,find,sshd -T,sysctl,visudo,chage,passwd,usermod) to audit and remediate system configurations. These operations are restricted to relevant security parameters and are consistent with the tool's primary purpose. - [EXTERNAL_DOWNLOADS]: The skill references several well-known and trusted external resources, including the CIS Benchmarks, GTFOBins, the Linux Kernel documentation, and OpenSSH project pages. These references are used solely for authoritative documentation and security best practices.
- [INDIRECT_PROMPT_INJECTION]: The skill processes system configuration files that could potentially contain untrusted data.
- Ingestion points: Reads from
/etc/passwd,/etc/shadow,/etc/ssh/sshd_config, and.ssh/authorized_keys(SKILL.md). - Boundary markers: None present for the file content ingestion.
- Capability inventory: Extensive system inspection and management capabilities including
awk,grep,find,sshd,sysctl, and administrative tools likeusermodandpasswd(SKILL.md). - Sanitization: The audit logic utilizes fixed regular expressions and specific field delimiters to parse system files, which mitigates the risk of command injection from file contents.
Audit Metadata