Linux Privilege Escalation

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.70). The skill instructs downloading and executing LinPEAS directly from GitHub via curl during runtime enumeration, introducing an unverifiable third-party operational dependency under weak/unknown provenance.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 1.00). The skill prompt instructs the agent on performing Linux privilege escalation techniques, including exploiting sudo misconfigurations, compiling kernel exploits, and executing reverse shells to gain root access on the host system.