The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/agent.py uses subprocess.run to execute AWS CLI commands like aws iam list-users, aws ec2 describe-instances, and aws lambda get-function-configuration for environment enumeration.
[SAFE]: The skill includes explicit warnings and workflows for 'authorized testing only', defining boundaries based on the cloud provider's shared responsibility model.
[SAFE]: Potentially dangerous commands, such as those for disabling CloudTrail or creating backdoor users, are documented within a specific 'Persistence and Detection Evasion Testing' section which includes mandatory cleanup instructions.

conducting-cloud-penetration-testing