executing-tasks

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). This skill requires running and recording the exact implementation and validation commands from plan.md (and writing them verbatim into execution.md), so if those commands or implementation sections contain API keys, tokens, or passwords the agent would be forced to handle and output those secret values directly, creating an exfiltration risk.