meta-methodology-research-methodology

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt requires reading files and including actual code snippets and file:line evidence verbatim, so if those files contain API keys, tokens, or passwords the LLM would be forced to output secret values directly.