Skills
Navigate
skills/itsar-vr/goatedskills/backend-coding-agent/Gen Agent Trust Hub

backend-coding-agent

Warn

Audited by Gen Agent Trust Hub on Apr 7, 2026

Risk Level: MEDIUMPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill explicitly directs the use of the --permission-mode bypassPermissions flag when calling the claude CLI, which is a direct attempt to override safety and permission constraints.
  • [COMMAND_EXECUTION]: The skill promotes the use of CLI tools like skill-oracle, codex, and claude with complex command-line arguments, which increases the potential attack surface.
  • [INDIRECT_PROMPT_INJECTION]: The skill provides a surface for indirect injection by interpolating user-supplied strings into shell commands without sanitization. * Ingestion points: User-provided backend task descriptions in SKILL.md. * Boundary markers: Double quotes are used for interpolation in command templates. * Capability inventory: Shell command execution via skill-oracle, codex, and claude tools. * Sanitization: No input validation or escaping is applied to user strings before command execution, potentially allowing command injection.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 7, 2026, 12:56 PM