gentleman-system

Warn

Audited by Gen Agent Trust Hub on Sep 19, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONPRIVILEGE_ESCALATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides a structured Go API for the agent to execute shell commands, including real-time logging and package management tasks.
  • [PRIVILEGE_ESCALATION]: The instructions include patterns for functions specifically designed to execute commands with root privileges via sudo (e.g., system.RunSudo).
  • [INDIRECT_PROMPT_INJECTION]: The skill defines a wide capability surface involving system-level operations without explicit instructions for sanitizing inputs that might originate from untrusted sources.
  • Ingestion points: Pattern definitions and code examples in SKILL.md.
  • Boundary markers: None present in the provided code snippets.
  • Capability inventory: Arbitrary shell execution, package installation (apt, pacman, brew, pkg), and file system operations (EnsureDir, CopyFile, CopyDir).
  • Sanitization: No evidence of input validation or command escaping in the provided execution patterns.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Sep 19, 2026, 12:22 AM
Security Audit — agent-trust-hub — gentleman-system