prompt-adapt

Fail

Audited by Gen Agent Trust Hub on Mar 22, 2026

Risk Level: HIGHCOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute a local Python script using a shell command: python3 {PROMPT_ENGINE_DIR}/scripts/search_prompts.py "SUBJECT" --model TARGET_MODEL --limit 3.
  • [COMMAND_EXECUTION]: User-controlled variables, specifically the 'SUBJECT' and 'TARGET_MODEL', are interpolated directly into the bash command string. Without strict sanitization, this allows an attacker to break out of the command context using shell metacharacters (e.g., ;, &, |, or `) to execute arbitrary code.
  • [COMMAND_EXECUTION]: The skill lacks instructions for validating or escaping user input before it is passed to the system shell, creating a direct path for system compromise via malicious prompts.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Mar 22, 2026, 09:14 PM