ctf-writeup

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt explicitly requires including the final flag verbatim in the writeup (unless redaction is requested) and instructs extracting flags from files/outputs, which forces the model to handle and output secret values directly.