examples-auto-run

Fail

Audited by Gen Agent Trust Hub on Aug 6, 2026

Risk Level: HIGHCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The script scripts/run.sh explicitly sets SHELL_AUTO_APPROVE=1 and APPLY_PATCH_AUTO_APPROVE=1. These environment variables are designed to bypass user confirmation for potentially dangerous operations, such as executing shell commands or applying file patches. This configuration overrides standard safety mechanisms that require a human-in-the-loop for sensitive system modifications.
  • [PROMPT_INJECTION]: The SKILL.md instructions define an 'auto mode' (EXAMPLES_INTERACTIVE_MODE=auto) that explicitly directs the agent to bypass user inputs and approvals during execution. While the skill instructs the agent to validate logs after the process completes, the underlying commands are executed without real-time human oversight or safety checkpoints.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Aug 6, 2026, 04:52 PM
Security Audit — agent-trust-hub — examples-auto-run