1password

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The skill largely promotes secure op:// references and env-based flows but also contains multiple examples that embed secrets inline (e.g., API_KEY=secret, export OP_SERVICE_ACCOUNT_TOKEN="ops_...", --from-literal=token=your-access-token, helm --set operator.token.value=...) which would require an LLM to include secret values verbatim in generated commands or manifests, posing an exfiltration risk.

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

• Potentially malicious external URL detected (high risk: 0.80). The Helm repo URL https://1password.github.io/connect-helm-charts is referenced as a runtime install step ("helm repo add" / "helm install") that fetches remote Helm charts (and ultimately container images like 1password/connect-api) which will deploy and execute remote code on your cluster, and the skill's native Kubernetes Operator integration relies on that external chart for operation.