container-security

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The report template and findings sections explicitly request "Evidence: " and call out "secrets in plaintext manifests" and base64-encoded secrets as findings, which can require the LLM to reproduce secret values verbatim in its output (exfiltration risk).

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.85). The required runtime workflow ingests the target repository’s Dockerfiles/Kubernetes manifests/Helm charts (outsider-authored configuration text) into the agent for evaluation, which is classic indirect prompt-injection exposure via YAML/Dockerfile string fields and comments.