evaluate-skill

Fail

Audited by Gen Agent Trust Hub on Apr 15, 2026

Risk Level: HIGHCOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses dynamic context injection syntax (!find $1/skills...) in the Context section of SKILL.md. This pattern executes shell commands at the moment the skill is loaded. Because the user-provided argument $1 is interpolated directly into the command string without sanitization, an attacker can provide a skill name containing shell metacharacters (such as ;, &, or $()) to execute arbitrary commands on the host system.
  • [COMMAND_EXECUTION]: The skill triggers the execution of local shell scripts (scripts/plugin-compliance-check.sh and evaluate-plugin/scripts/aggregate_benchmark.sh) using the bash utility. While these are local references, the safety of this operation is contingent on the integrity and sanitization logic within those external scripts.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Apr 15, 2026, 10:01 AM