linux-privilege-escalation

Fail

Audited by Gen Agent Trust Hub on Jun 14, 2026

Risk Level: HIGHCOMMAND_EXECUTIONDATA_EXFILTRATIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCREDENTIALS_UNSAFE
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill instructs the user to download scripts and source code from external URLs (GitHub and arbitrary IP addresses) and execute them on the target system (e.g., curl ... linpeas.sh; ./linpeas.sh).
  • [COMMAND_EXECUTION]: Contains numerous commands for hijacking system control, including compiling C exploits with gcc, modifying critical system files like /etc/passwd and /etc/crontab, and manipulating environment variables like PATH for hijacking.
  • [DATA_EXFILTRATION]: Provides templates for establishing persistent reverse shell connections to external servers using various protocols and languages (Bash, Python, Perl, and Netcat).
  • [CREDENTIALS_UNSAFE]: Details methods for accessing and reading the /etc/shadow file, which contains hashed user passwords, to enable offline credential cracking.
  • [EXTERNAL_DOWNLOADS]: Routinely employs curl and wget to fetch external enumeration scripts, exploit source code, and binaries from non-standard or user-defined remote locations.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Jun 14, 2026, 11:29 AM
Security Audit — agent-trust-hub — linux-privilege-escalation