vet

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill mandates including file paths and actual code snippets as evidence for findings (including "Secrets in code or config"), which forces the model to reproduce any secret values found in the repository verbatim in its output, creating an exfiltration risk.

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

• Third-party content exposure detected (medium risk: 0.30). The runtime workflow for /vet reads first-party project files in a user-supplied codebase (README/config/code) as part of mapping and investigation, so outsider-authored free text can be ingested indirectly via those repository documents and source files.