Cloud Penetration Testing

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). This skill contains multiple instructions that extract, print, or pass secrets on command lines or in variables (e.g., --secret_access_key, converting $spn.Secret to an unsecure string, --password, importing/storing stolen token files), which would require an LLM to handle or reproduce secret values verbatim.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). The content contains explicit, actionable instructions for credential theft, secret exfiltration, creating persistent backdoors (service principals and access keys), password spraying, and metadata/instance-token harvesting that enable unauthorized remote access and data exfiltration.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill's Required Tools installation fetches and executes remote installers at runtime (curl "https://awscli.amazonaws.com/awscli-exe-linux-x86_64.zip" then ./aws/install and curl https://sdk.cloud.google.com | bash), which downloads and runs remote code, creating a high-confidence runtime dependency.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 1.00). The skill includes explicit sudo commands and instructions that install or modify software/configuration on the host (e.g., "sudo ./aws/install", "curl ... | bash", and "sudo cp" / "sudo find"), which push an agent to obtain elevated privileges and change the machine state.