art-to-web-analysis

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The prompt mandates reproducing exact visible text character-by-character (including punctuation and whitespace), so if a provided image contains API keys, tokens, or passwords the LLM would be required to output those secrets verbatim.