gcp-monitoring

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The prompt is mostly safe with placeholders and CLI/SDK usage, but it includes an explicit insecure example that passes a PagerDuty service_key as a command-line label (service_key=PD_SERVICE_KEY), which would require embedding a secret verbatim into generated commands—an exfiltration risk.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill includes runtime installation commands that fetch and execute remote code — notably curling and then running the script at https://dl.google.com/cloudagents/add-google-cloud-ops-agent-repo.sh (curl ...; sudo bash ...) and pip installs (pip install google-cloud-trace / google-cloud-profiler) which download and install remote packages, so external content is executed as part of the skill.

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 tells the agent to run sudo to install the Google Cloud Ops Agent, write/modify /etc/google-cloud-ops-agent/config.yaml, and restart a systemctl service — explicit modification of system files and services requiring elevated privileges, so it can change the machine state.