fuck-slop

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The skill explicitly instructs the agent to produce a findings list and "report the findings ... (category, count, worst example)" including the offending line/sentence, which would require the LLM to reproduce input text snippets verbatim — so if those snippets contain secrets (API keys, tokens, passwords), the model is required to output them.