review-implementation

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill requires the agent to inspect local/uncommitted files and "include verification results with the exact commands that were discovered and run," which would force verbatim output of any secrets embedded in those commands or files (e.g., API keys passed on the CLI or in scripts).