implementing-beyondcorp-zero-trust-access-model

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The prompt instructs embedding sensitive values (e.g., --oauth2-client-secret, --oauth2-client-id and similar placeholders) directly into gcloud command-line flags, which would require the LLM to include secret values verbatim in outputs/commands and creates an exfiltration risk.