aws-fundamentals

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt includes hardcoded credentials (e.g., the RDS example: --master-user-password secretpassword) and CLI examples that embed passwords directly, which forces the agent to output secret values verbatim and poses an exfiltration risk.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The EC2 user-data script explicitly runs "aws s3 cp s3://my-bucket/app.zip /home/ec2-user/ && unzip app.zip" and then starts the app with gunicorn, meaning s3://my-bucket/app.zip is fetched at runtime and its contents are executed on the instance.